Attachment 36379 Details for Bug 28624 – stderr with RADEON_DEBUG=fp

stderr with RADEON_DEBUG=fp

etqw_fp.txt (text/plain), 700.47 KB, created by almos on 2010-06-20 06:44:50 UTC

(hide)

Description:

Filename:

MIME Type:

Creator: almos

Created: 2010-06-20 06:44:50 UTC

Size: 700.47 KB

patch

obsolete

>debug_get_option: GALLIUM_TRACE = (null)
>debug_get_bool_option: GALLIUM_RBUG = FALSE
>radeon: Successfully grabbed chipset info from kernel!
>radeon: DRM version: 2.0.0 ID: 0x4152 GB: 1 Z: 1
>radeon: GART size: 61 MB VRAM size: 128 MB
>debug_get_flags_option: RADEON_DEBUG = 0x2 (fp)
>debug_get_bool_option: RADEON_NO_TCL = FALSE
>Mesa: CPU vendor: GenuineIntel
>Mesa: CPU name:               Intel(R) Pentium(R) 4 CPU 2.80GHz
>Mesa: Mesa 7.9-devel DEBUG build Jun 10 2010 23:16:01
>Mesa warning: software DXTn compression/decompression available
>debug_get_bool_option: MESA_MVP_DP4 = FALSE
>debug_get_flags_option: ST_DEBUG = 0x0
>debug_get_bool_option: DRAW_FSE = FALSE
>debug_get_bool_option: DRAW_NO_FSE = FALSE
>debug_get_bool_option: GALLIUM_DUMP_VS = FALSE
>r300: Initial fragment program
>FRAG
>DCL IN[0], GENERIC[0], LINEAR
>DCL OUT[0], COLOR
>DCL SAMP[0]
>  0: TEX OUT[0], IN[0], SAMP[0], 2D
>  1: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: TEX output[0], input[0], 2D[0];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: TEX output[0], input[0], 2D[0];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: TEX output[0], input[0], 2D[0];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: TEX output[0], input[0], 2D[0];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: TEX temp[1], input[0], 2D[0];
>  1: MOV output[0], temp[1];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: TEX temp[1], input[0].xy__, 2D[0];
>  1: MOV output[0], temp[1];
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: TEX temp[1], input[0].xy__, 2D[0];
>  1: MOV output[0], temp[1];
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: TEX temp[1], input[0].xy__, 2D[0];
>  1: MOV output[0], temp[1];
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: TEX temp[1], input[0].xy__, 2D[0];
>  1: src0.xyz = temp[1], src0.w = temp[1]
>     MAD color[0].xyz, src0.xyz, src0.111, src0.000
>     MAD color[0].w, src0.w, src0.1, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[1], input[0].xy__, 2D[0];
>  2: src0.xyz = temp[1], src0.w = temp[1]
>     MAD color[0].xyz, src0.xyz, src0.111, src0.000
>     MAD color[0].w, src0.w, src0.1, src0.0
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[0], temp[0].xy__, 2D[0];
>  2: src0.xyz = temp[0], src0.w = temp[0]
>     MAD color[0].xyz, src0.xyz, src0.111, src0.000
>     MAD color[0].w, src0.w, src0.1, src0.0
>pc=0*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 0, tex_end: 0  (code_addr: 00400000)
>  TEX:
>    TEX t0, t0, texture[0]   (00008000)
>  0: xyz:  t0  t0  t0 -> o0.xyz               (1c000000)
>       w:  t0  t0  t0 -> o0.w                 (01000000)
>     xyz:   t0.xyz      1.0      0.0    op: 00050a80
>       w:     t0.w      1.0      0.0    op: 00040889
>r300: Initial fragment program
>FRAG
>DCL IN[0], COLOR, LINEAR
>DCL OUT[0], COLOR
>  0: MOV OUT[0], IN[0]
>  1: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: src0.xyz = input[0], src0.w = input[0]
>     MAD color[0].xyz, src0.xyz, src0.111, src0.000
>     MAD color[0].w, src0.w, src0.1, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: src0.xyz = input[0], src0.w = input[0]
>     MAD color[0].xyz, src0.xyz, src0.111, src0.000
>     MAD color[0].w, src0.w, src0.1, src0.0
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: src0.xyz = temp[0], src0.w = temp[0]
>     MAD color[0].xyz, src0.xyz, src0.111, src0.000
>     MAD color[0].w, src0.w, src0.1, src0.0
>pc=1*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 0, tex_end: 0  (code_addr: 00400000)
>  0: xyz:  t0  t0  t0 -> o0.xyz               (1c000000)
>       w:  t0  t0  t0 -> o0.w                 (01000000)
>     xyz:   t0.xyz      1.0      0.0    op: 00050a80
>       w:     t0.w      1.0      0.0    op: 00040889
>r300: Initial fragment program
>FRAG
>DCL IN[0], COLOR, PERSPECTIVE
>DCL OUT[0], COLOR
>  0: MOV OUT[0], IN[0]
>  1: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: src0.xyz = input[0], src0.w = input[0]
>     MAD color[0].xyz, src0.xyz, src0.111, src0.000
>     MAD color[0].w, src0.w, src0.1, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: src0.xyz = input[0], src0.w = input[0]
>     MAD color[0].xyz, src0.xyz, src0.111, src0.000
>     MAD color[0].w, src0.w, src0.1, src0.0
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: src0.xyz = temp[0], src0.w = temp[0]
>     MAD color[0].xyz, src0.xyz, src0.111, src0.000
>     MAD color[0].w, src0.w, src0.1, src0.0
>pc=2*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 0, tex_end: 0  (code_addr: 00400000)
>  0: xyz:  t0  t0  t0 -> o0.xyz               (1c000000)
>       w:  t0  t0  t0 -> o0.w                 (01000000)
>     xyz:   t0.xyz      1.0      0.0    op: 00050a80
>       w:     t0.w      1.0      0.0    op: 00040889
>r300: Initial fragment program
>FRAG
>DCL IN[0], COLOR, LINEAR
>DCL IN[1], GENERIC[0], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL TEMP[0]
>  0: TEX TEMP[0], IN[1], SAMP[0], 2D
>  1: MUL OUT[0], TEMP[0], IN[0]
>  2: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: MUL output[0], temp[0], input[0];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: MUL output[0], temp[0], input[0];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: MUL output[0], temp[0], input[0];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: MUL output[0], temp[0], input[0];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: MUL output[0], temp[0], input[0];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: TEX temp[0], input[1].xy__, 2D[0];
>  1: MUL output[0], temp[0], input[0];
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: TEX temp[0], input[1].xy__, 2D[0];
>  1: MUL output[0], temp[0], input[0];
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: TEX temp[0], input[1].xy__, 2D[0];
>  1: MUL output[0], temp[0], input[0];
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: TEX temp[0], input[1].xy__, 2D[0];
>  1: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = input[0], src1.w = input[0]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[0], input[1].xy__, 2D[0];
>  2: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = input[0], src1.w = input[0]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[1], temp[1].xy__, 2D[0];
>  2: src0.xyz = temp[1], src0.w = temp[1], src1.xyz = temp[0], src1.w = temp[0]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>pc=3*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 0, tex_end: 0  (code_addr: 00400000)
>  TEX:
>    TEX t1, t1, texture[0]   (00008041)
>  0: xyz:  t1  t0  t0 -> o0.xyz               (1c000001)
>       w:  t1  t0  t0 -> o0.w                 (01000001)
>     xyz:   t1.xyz   t0.xyz      0.0    op: 00050200
>       w:     t1.w     t0.w      0.0    op: 00040509
>r300: Initial fragment program
>FRAG
>DCL OUT[0], COLOR
>DCL CONST[0]
>  0: MOV OUT[0], CONST[0]
>  1: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: MOV output[0], const[0];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: MOV output[0], const[0];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: MOV output[0], const[0];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: MOV output[0], const[0];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: MOV output[0], const[0];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: MOV output[0], const[0];
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: MOV output[0], const[0];
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: MOV output[0], const[0];
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: src0.xyz = const[0], src0.w = const[0]
>     MAD color[0].xyz, src0.xyz, src0.111, src0.000
>     MAD color[0].w, src0.w, src0.1, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: src0.xyz = const[0], src0.w = const[0]
>     MAD color[0].xyz, src0.xyz, src0.111, src0.000
>     MAD color[0].w, src0.w, src0.1, src0.0
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: src0.xyz = const[0], src0.w = const[0]
>     MAD color[0].xyz, src0.xyz, src0.111, src0.000
>     MAD color[0].w, src0.w, src0.1, src0.0
>pc=4*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 0, tex_end: 0  (code_addr: 00400000)
>  0: xyz:  c0  t0  t0 -> o0.xyz               (1c000020)
>       w:  c0  t0  t0 -> o0.w                 (01000020)
>     xyz:   c0.xyz      1.0      0.0    op: 00050a80
>       w:     c0.w      1.0      0.0    op: 00040889
>r300: Initial fragment program
>FRAG
>DCL OUT[0], COLOR
>IMM FLT32 {    0.0000,     0.0000,     0.0000,     0.0000}
>  0: MOV OUT[0], IMM[0].xxxx
>  1: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0000;
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0000;
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0000;
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0000;
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0000;
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0000;
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0000;
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0000;
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: 
>     MAD color[0].xyz, src0.000, src0.111, src0.000
>     MAD color[0].w, src0.0, src0.1, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: 
>     MAD color[0].xyz, src0.000, src0.111, src0.000
>     MAD color[0].w, src0.0, src0.1, src0.0
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: 
>     MAD color[0].xyz, src0.000, src0.111, src0.000
>     MAD color[0].w, src0.0, src0.1, src0.0
>pc=5*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 0, tex_end: 0  (code_addr: 00400000)
>  0: xyz:  t0  t0  t0 -> o0.xyz               (1c000000)
>       w:  t0  t0  t0 -> o0.w                 (01000000)
>     xyz:      0.0      1.0      0.0    op: 00050a94
>       w:      0.0      1.0      0.0    op: 00040890
>r300: Initial fragment program
>FRAG
>DCL IN[0], COLOR, LINEAR
>DCL IN[1], GENERIC[2], PERSPECTIVE
>DCL IN[2], GENERIC[4], PERSPECTIVE
>DCL IN[3], GENERIC[5], PERSPECTIVE
>DCL IN[4], GENERIC[6], PERSPECTIVE
>DCL IN[5], GENERIC[7], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL SAMP[1]
>DCL SAMP[2]
>DCL SAMP[3]
>DCL SAMP[4]
>DCL SAMP[5]
>DCL CONST[3..5]
>DCL TEMP[0..11]
>IMM FLT32 {    0.0010,     0.0000,     1.0000,     0.1000}
>IMM FLT32 {    2.0000,    -1.0000,     8.0000,     0.0000}
>  0: TEX TEMP[0], IN[2], SAMP[0], 2D
>  1: SUB TEMP[1].w, TEMP[0].wwww, IMM[0].xxxx
>  2: KIL TEMP[1].wwww
>  3: TEX TEMP[2], IN[2], SAMP[1], 2D
>  4: MOV TEMP[2].x, TEMP[2].wwww
>  5: MAD TEMP[2].xy, TEMP[2], IMM[1].xxxx, IMM[1].yyyy
>  6: MOV TEMP[2].z, IMM[0].yyyy
>  7: DP3 TEMP[3].x, TEMP[2], TEMP[2]
>  8: ADD TEMP[3].x, IMM[0].zzzz, -TEMP[3].xxxx
>  9: RSQ TEMP[3].x, TEMP[3].xxxx
> 10: RCP TEMP[2].z, TEMP[3].xxxx
> 11: DP3 TEMP[4].x, IN[3], TEMP[2]
> 12: DP3 TEMP[4].y, IN[4], TEMP[2]
> 13: DP3 TEMP[4].z, IN[5], TEMP[2]
> 14: TEX TEMP[5], TEMP[4], SAMP[2], CUBE
> 15: MUL TEMP[5], TEMP[5], CONST[3]
> 16: MUL TEMP[5], TEMP[5], IMM[1].xxxx
> 17: MUL TEMP[6], TEMP[0], CONST[4]
> 18: MUL TEMP[6], TEMP[5], TEMP[6]
> 19: DP3 TEMP[3], IN[1], TEMP[4]
> 20: ADD TEMP[3], TEMP[3], TEMP[3]
> 21: MAD TEMP[7], TEMP[4], TEMP[3].xxxx, -IN[1]
> 22: TEX TEMP[8], TEMP[7], SAMP[3], CUBE
> 23: TEX TEMP[9], IN[2], SAMP[4], 2D
> 24: ADD TEMP[1], TEMP[9], TEMP[9]
> 25: MAD TEMP[6], TEMP[1], TEMP[8], TEMP[6]
> 26: MUL TEMP[3], TEMP[6], IN[0]
> 27: MOV OUT[0].w, TEMP[0].wwww
> 28: DP3 TEMP[1], IN[1], TEMP[4]
> 29: ADD TEMP[1], IMM[0].zzzz, -TEMP[1]
> 30: MUL TEMP[1], TEMP[1], TEMP[1]
> 31: MUL TEMP[1], TEMP[1], TEMP[1]
> 32: MUL TEMP[1], TEMP[1], CONST[5]
> 33: MUL TEMP[10], TEMP[1], TEMP[5]
> 34: MAD TEMP[10], TEMP[1], IMM[0].wwww, TEMP[10]
> 35: TEX TEMP[11], IN[2], SAMP[5], 2D
> 36: MUL_SAT TEMP[5], TEMP[5], IMM[1].zzzz
> 37: MAD TEMP[1], TEMP[11], -TEMP[5].xxxx, TEMP[11]
> 38: MAX TEMP[1], TEMP[1], IMM[0].yyyy
> 39: ADD TEMP[10], TEMP[1], TEMP[10]
> 40: ADD OUT[0].xyz, TEMP[3], TEMP[10]
> 41: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: TEX temp[0], input[2], 2D[0];
>  1: SUB temp[1].w, temp[0].wwww, const[6].xxxx;
>  2: KIL temp[1].wwww;
>  3: TEX temp[2], input[2], 2D[1];
>  4: MOV temp[2].x, temp[2].wwww;
>  5: MAD temp[2].xy, temp[2], const[7].xxxx, const[7].yyyy;
>  6: MOV temp[2].z, const[6].yyyy;
>  7: DP3 temp[3].x, temp[2], temp[2];
>  8: ADD temp[3].x, const[6].zzzz, -temp[3].xxxx;
>  9: RSQ temp[3].x, temp[3].xxxx;
> 10: RCP temp[2].z, temp[3].xxxx;
> 11: DP3 temp[4].x, input[3], temp[2];
> 12: DP3 temp[4].y, input[4], temp[2];
> 13: DP3 temp[4].z, input[5], temp[2];
> 14: TEX temp[5], temp[4], CUBE[2];
> 15: MUL temp[5], temp[5], const[3];
> 16: MUL temp[5], temp[5], const[7].xxxx;
> 17: MUL temp[6], temp[0], const[4];
> 18: MUL temp[6], temp[5], temp[6];
> 19: DP3 temp[3], input[1], temp[4];
> 20: ADD temp[3], temp[3], temp[3];
> 21: MAD temp[7], temp[4], temp[3].xxxx, -input[1];
> 22: TEX temp[8], temp[7], CUBE[3];
> 23: TEX temp[9], input[2], 2D[4];
> 24: ADD temp[1], temp[9], temp[9];
> 25: MAD temp[6], temp[1], temp[8], temp[6];
> 26: MUL temp[3], temp[6], input[0];
> 27: MOV output[0].w, temp[0].wwww;
> 28: DP3 temp[1], input[1], temp[4];
> 29: ADD temp[1], const[6].zzzz, -temp[1];
> 30: MUL temp[1], temp[1], temp[1];
> 31: MUL temp[1], temp[1], temp[1];
> 32: MUL temp[1], temp[1], const[5];
> 33: MUL temp[10], temp[1], temp[5];
> 34: MAD temp[10], temp[1], const[6].wwww, temp[10];
> 35: TEX temp[11], input[2], 2D[5];
> 36: MUL_SAT temp[5], temp[5], const[7].zzzz;
> 37: MAD temp[1], temp[11], -temp[5].xxxx, temp[11];
> 38: MAX temp[1], temp[1], const[6].yyyy;
> 39: ADD temp[10], temp[1], temp[10];
> 40: ADD output[0].xyz, temp[3], temp[10];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: TEX temp[0], input[2], 2D[0];
>  1: SUB temp[1].w, temp[0].wwww, const[6].xxxx;
>  2: KIL temp[1].wwww;
>  3: TEX temp[2], input[2], 2D[1];
>  4: MOV temp[2].x, temp[2].wwww;
>  5: MAD temp[2].xy, temp[2], const[7].xxxx, const[7].yyyy;
>  6: MOV temp[2].z, const[6].yyyy;
>  7: DP3 temp[3].x, temp[2], temp[2];
>  8: ADD temp[3].x, const[6].zzzz, -temp[3].xxxx;
>  9: RSQ temp[3].x, temp[3].xxxx;
> 10: RCP temp[2].z, temp[3].xxxx;
> 11: DP3 temp[4].x, input[3], temp[2];
> 12: DP3 temp[4].y, input[4], temp[2];
> 13: DP3 temp[4].z, input[5], temp[2];
> 14: TEX temp[5], temp[4], CUBE[2];
> 15: MUL temp[5], temp[5], const[3];
> 16: MUL temp[5], temp[5], const[7].xxxx;
> 17: MUL temp[6], temp[0], const[4];
> 18: MUL temp[6], temp[5], temp[6];
> 19: DP3 temp[3], input[1], temp[4];
> 20: ADD temp[3], temp[3], temp[3];
> 21: MAD temp[7], temp[4], temp[3].xxxx, -input[1];
> 22: TEX temp[8], temp[7], CUBE[3];
> 23: TEX temp[9], input[2], 2D[4];
> 24: ADD temp[1], temp[9], temp[9];
> 25: MAD temp[6], temp[1], temp[8], temp[6];
> 26: MUL temp[3], temp[6], input[0];
> 27: MOV output[0].w, temp[0].wwww;
> 28: DP3 temp[1], input[1], temp[4];
> 29: ADD temp[1], const[6].zzzz, -temp[1];
> 30: MUL temp[1], temp[1], temp[1];
> 31: MUL temp[1], temp[1], temp[1];
> 32: MUL temp[1], temp[1], const[5];
> 33: MUL temp[10], temp[1], temp[5];
> 34: MAD temp[10], temp[1], const[6].wwww, temp[10];
> 35: TEX temp[11], input[2], 2D[5];
> 36: MUL_SAT temp[5], temp[5], const[7].zzzz;
> 37: MAD temp[1], temp[11], -temp[5].xxxx, temp[11];
> 38: MAX temp[1], temp[1], const[6].yyyy;
> 39: ADD temp[10], temp[1], temp[10];
> 40: ADD output[0].xyz, temp[3], temp[10];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: TEX temp[0], input[2], 2D[0];
>  1: SUB temp[1].w, temp[0].wwww, const[6].xxxx;
>  2: KIL temp[1].wwww;
>  3: TEX temp[2], input[2], 2D[1];
>  4: MOV temp[2].x, temp[2].wwww;
>  5: MAD temp[2].xy, temp[2], const[7].xxxx, const[7].yyyy;
>  6: MOV temp[2].z, const[6].yyyy;
>  7: DP3 temp[3].x, temp[2], temp[2];
>  8: ADD temp[3].x, const[6].zzzz, -temp[3].xxxx;
>  9: RSQ temp[3].x, temp[3].xxxx;
> 10: RCP temp[2].z, temp[3].xxxx;
> 11: DP3 temp[4].x, input[3], temp[2];
> 12: DP3 temp[4].y, input[4], temp[2];
> 13: DP3 temp[4].z, input[5], temp[2];
> 14: TEX temp[5], temp[4], CUBE[2];
> 15: MUL temp[5], temp[5], const[3];
> 16: MUL temp[5], temp[5], const[7].xxxx;
> 17: MUL temp[6], temp[0], const[4];
> 18: MUL temp[6], temp[5], temp[6];
> 19: DP3 temp[3], input[1], temp[4];
> 20: ADD temp[3], temp[3], temp[3];
> 21: MAD temp[7], temp[4], temp[3].xxxx, -input[1];
> 22: TEX temp[8], temp[7], CUBE[3];
> 23: TEX temp[9], input[2], 2D[4];
> 24: ADD temp[1], temp[9], temp[9];
> 25: MAD temp[6], temp[1], temp[8], temp[6];
> 26: MUL temp[3], temp[6], input[0];
> 27: MOV output[0].w, temp[0].wwww;
> 28: DP3 temp[1], input[1], temp[4];
> 29: ADD temp[1], const[6].zzzz, -temp[1];
> 30: MUL temp[1], temp[1], temp[1];
> 31: MUL temp[1], temp[1], temp[1];
> 32: MUL temp[1], temp[1], const[5];
> 33: MUL temp[10], temp[1], temp[5];
> 34: MAD temp[10], temp[1], const[6].wwww, temp[10];
> 35: TEX temp[11], input[2], 2D[5];
> 36: MUL_SAT temp[5], temp[5], const[7].zzzz;
> 37: MAD temp[1], temp[11], -temp[5].xxxx, temp[11];
> 38: MAX temp[1], temp[1], const[6].yyyy;
> 39: ADD temp[10], temp[1], temp[10];
> 40: ADD output[0].xyz, temp[3], temp[10];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: TEX temp[0], input[2], 2D[0];
>  1: ADD temp[1].w, temp[0].wwww, -const[6].xxxx;
>  2: KIL temp[1].wwww;
>  3: TEX temp[2], input[2], 2D[1];
>  4: MOV temp[2].x, temp[2].wwww;
>  5: MAD temp[2].xy, temp[2], const[7].xxxx, const[7].yyyy;
>  6: MOV temp[2].z, const[6].yyyy;
>  7: DP3 temp[3].x, temp[2], temp[2];
>  8: ADD temp[3].x, const[6].zzzz, -temp[3].xxxx;
>  9: RSQ temp[3].x, |temp[3].xxxx|;
> 10: RCP temp[2].z, temp[3].xxxx;
> 11: DP3 temp[4].x, input[3], temp[2];
> 12: DP3 temp[4].y, input[4], temp[2];
> 13: DP3 temp[4].z, input[5], temp[2];
> 14: TEX temp[5], temp[4], CUBE[2];
> 15: MUL temp[5], temp[5], const[3];
> 16: MUL temp[5], temp[5], const[7].xxxx;
> 17: MUL temp[6], temp[0], const[4];
> 18: MUL temp[6], temp[5], temp[6];
> 19: DP3 temp[3], input[1], temp[4];
> 20: ADD temp[3], temp[3], temp[3];
> 21: MAD temp[7], temp[4], temp[3].xxxx, -input[1];
> 22: TEX temp[8], temp[7], CUBE[3];
> 23: TEX temp[9], input[2], 2D[4];
> 24: ADD temp[1], temp[9], temp[9];
> 25: MAD temp[6], temp[1], temp[8], temp[6];
> 26: MUL temp[3], temp[6], input[0];
> 27: MOV output[0].w, temp[0].wwww;
> 28: DP3 temp[1], input[1], temp[4];
> 29: ADD temp[1], const[6].zzzz, -temp[1];
> 30: MUL temp[1], temp[1], temp[1];
> 31: MUL temp[1], temp[1], temp[1];
> 32: MUL temp[1], temp[1], const[5];
> 33: MUL temp[10], temp[1], temp[5];
> 34: MAD temp[10], temp[1], const[6].wwww, temp[10];
> 35: TEX temp[11], input[2], 2D[5];
> 36: MUL_SAT temp[5], temp[5], const[7].zzzz;
> 37: MAD temp[1], temp[11], -temp[5].xxxx, temp[11];
> 38: MAX temp[1], temp[1], const[6].yyyy;
> 39: ADD temp[10], temp[1], temp[10];
> 40: ADD output[0].xyz, temp[3], temp[10];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: TEX temp[0], input[2], 2D[0];
>  1: ADD temp[1].w, temp[0].wwww, -const[6].xxxx;
>  2: KIL temp[1].wwww;
>  3: TEX temp[2], input[2], 2D[1];
>  4: MOV temp[2].x, temp[2].wwww;
>  5: MAD temp[2].xy, temp[2], const[7].xxxx, const[7].yyyy;
>  6: MOV temp[2].z, const[6].yyyy;
>  7: DP3 temp[3].x, temp[2], temp[2];
>  8: ADD temp[3].x, const[6].zzzz, -temp[3].xxxx;
>  9: RSQ temp[3].x, |temp[3].xxxx|;
> 10: RCP temp[2].z, temp[3].xxxx;
> 11: DP3 temp[4].x, input[3], temp[2];
> 12: DP3 temp[4].y, input[4], temp[2];
> 13: DP3 temp[4].z, input[5], temp[2];
> 14: TEX temp[5], temp[4], CUBE[2];
> 15: MUL temp[5], temp[5], const[3];
> 16: MUL temp[5], temp[5], const[7].xxxx;
> 17: MUL temp[6], temp[0], const[4];
> 18: MUL temp[6], temp[5], temp[6];
> 19: DP3 temp[3], input[1], temp[4];
> 20: ADD temp[3], temp[3], temp[3];
> 21: MAD temp[7], temp[4], temp[3].xxxx, -input[1];
> 22: TEX temp[8], temp[7], CUBE[3];
> 23: TEX temp[9], input[2], 2D[4];
> 24: ADD temp[1], temp[9], temp[9];
> 25: MAD temp[6], temp[1], temp[8], temp[6];
> 26: MUL temp[3], temp[6], input[0];
> 27: MOV output[0].w, temp[0].wwww;
> 28: DP3 temp[1], input[1], temp[4];
> 29: ADD temp[1], const[6].zzzz, -temp[1];
> 30: MUL temp[1], temp[1], temp[1];
> 31: MUL temp[1], temp[1], temp[1];
> 32: MUL temp[1], temp[1], const[5];
> 33: MUL temp[10], temp[1], temp[5];
> 34: MAD temp[10], temp[1], const[6].wwww, temp[10];
> 35: TEX temp[11], input[2], 2D[5];
> 36: MUL_SAT temp[5], temp[5], const[7].zzzz;
> 37: MAD temp[1], temp[11], -temp[5].xxxx, temp[11];
> 38: MAX temp[1], temp[1], const[6].yyyy;
> 39: ADD temp[10], temp[1], temp[10];
> 40: ADD output[0].xyz, temp[3], temp[10];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: TEX temp[0], input[2].xy__, 2D[0];
>  1: ADD temp[1].w, temp[0].___w, -const[6].___x;
>  2: KIL temp[1].wwww;
>  3: TEX temp[2].yw, input[2].xy__, 2D[1];
>  4: MOV temp[2].x, temp[2].w___;
>  5: MAD temp[2].xy, temp[2].xy__, const[7].xx__, const[7].yy__;
>  6: MOV temp[2].z, const[6].__y_;
>  7: DP3 temp[3].x, temp[2].xyz_, temp[2].xyz_;
>  8: ADD temp[3].x, const[6].z___, -temp[3].x___;
>  9: RSQ temp[3].x, |temp[3].x___|;
> 10: RCP temp[2].z, temp[3].x___;
> 11: DP3 temp[4].x, input[3].xyz_, temp[2].xyz_;
> 12: DP3 temp[4].y, input[4].xyz_, temp[2].xyz_;
> 13: DP3 temp[4].z, input[5].xyz_, temp[2].xyz_;
> 14: TEX temp[5].xyz, temp[4].xyz_, CUBE[2];
> 15: MUL temp[5].xyz, temp[5].xyz_, const[3].xyz_;
> 16: MUL temp[5].xyz, temp[5].xyz_, const[7].xxx_;
> 17: MUL temp[6].xyz, temp[0].xyz_, const[4].xyz_;
> 18: MUL temp[6].xyz, temp[5].xyz_, temp[6].xyz_;
> 19: DP3 temp[3].x, input[1].xyz_, temp[4].xyz_;
> 20: ADD temp[3].x, temp[3].x___, temp[3].x___;
> 21: MAD temp[7].xyz, temp[4].xyz_, temp[3].xxx_, -input[1].xyz_;
> 22: TEX temp[8].xyz, temp[7].xyz_, CUBE[3];
> 23: TEX temp[9].xyz, input[2].xy__, 2D[4];
> 24: ADD temp[1].xyz, temp[9].xyz_, temp[9].xyz_;
> 25: MAD temp[6].xyz, temp[1].xyz_, temp[8].xyz_, temp[6].xyz_;
> 26: MUL temp[3].xyz, temp[6].xyz_, input[0].xyz_;
> 27: MOV output[0].w, temp[0].___w;
> 28: DP3 temp[1].xyz, input[1].xyz_, temp[4].xyz_;
> 29: ADD temp[1].xyz, const[6].zzz_, -temp[1].xyz_;
> 30: MUL temp[1].xyz, temp[1].xyz_, temp[1].xyz_;
> 31: MUL temp[1].xyz, temp[1].xyz_, temp[1].xyz_;
> 32: MUL temp[1].xyz, temp[1].xyz_, const[5].xyz_;
> 33: MUL temp[10].xyz, temp[1].xyz_, temp[5].xyz_;
> 34: MAD temp[10].xyz, temp[1].xyz_, const[6].www_, temp[10].xyz_;
> 35: TEX temp[11].xyz, input[2].xy__, 2D[5];
> 36: MUL_SAT temp[5].x, temp[5].x___, const[7].z___;
> 37: MAD temp[1].xyz, temp[11].xyz_, -temp[5].xxx_, temp[11].xyz_;
> 38: MAX temp[1].xyz, temp[1].xyz_, const[6].yyy_;
> 39: ADD temp[10].xyz, temp[1].xyz_, temp[10].xyz_;
> 40: ADD output[0].xyz, temp[3].xyz_, temp[10].xyz_;
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: TEX temp[0], input[2].xy__, 2D[0];
>  1: ADD temp[1].w, temp[0].___w, -const[6].___x;
>  2: KIL temp[1].wwww;
>  3: TEX temp[2].yw, input[2].xy__, 2D[1];
>  4: MOV temp[2].x, temp[2].w___;
>  5: MAD temp[2].xy, temp[2].xy__, const[7].xx__, none.-1-1__;
>  6: MOV temp[2].z, none.__0_;
>  7: DP3 temp[3].x, temp[2].xyz_, temp[2].xyz_;
>  8: ADD temp[3].x, none.1___, -temp[3].x___;
>  9: RSQ temp[3].x, |temp[3].x___|;
> 10: RCP temp[2].z, temp[3].x___;
> 11: DP3 temp[4].x, input[3].xyz_, temp[2].xyz_;
> 12: DP3 temp[4].y, input[4].xyz_, temp[2].xyz_;
> 13: DP3 temp[4].z, input[5].xyz_, temp[2].xyz_;
> 14: TEX temp[5].xyz, temp[4].xyz_, CUBE[2];
> 15: MUL temp[5].xyz, temp[5].xyz_, const[3].xyz_;
> 16: MUL temp[5].xyz, temp[5].xyz_, const[7].xxx_;
> 17: MUL temp[6].xyz, temp[0].xyz_, const[4].xyz_;
> 18: MUL temp[6].xyz, temp[5].xyz_, temp[6].xyz_;
> 19: DP3 temp[3].x, input[1].xyz_, temp[4].xyz_;
> 20: ADD temp[3].x, temp[3].x___, temp[3].x___;
> 21: MAD temp[7].xyz, temp[4].xyz_, temp[3].xxx_, -input[1].xyz_;
> 22: TEX temp[8].xyz, temp[7].xyz_, CUBE[3];
> 23: TEX temp[9].xyz, input[2].xy__, 2D[4];
> 24: ADD temp[1].xyz, temp[9].xyz_, temp[9].xyz_;
> 25: MAD temp[6].xyz, temp[1].xyz_, temp[8].xyz_, temp[6].xyz_;
> 26: MUL temp[3].xyz, temp[6].xyz_, input[0].xyz_;
> 27: MOV output[0].w, temp[0].___w;
> 28: DP3 temp[1].xyz, input[1].xyz_, temp[4].xyz_;
> 29: ADD temp[1].xyz, none.111_, -temp[1].xyz_;
> 30: MUL temp[1].xyz, temp[1].xyz_, temp[1].xyz_;
> 31: MUL temp[1].xyz, temp[1].xyz_, temp[1].xyz_;
> 32: MUL temp[1].xyz, temp[1].xyz_, const[5].xyz_;
> 33: MUL temp[10].xyz, temp[1].xyz_, temp[5].xyz_;
> 34: MAD temp[10].xyz, temp[1].xyz_, const[6].www_, temp[10].xyz_;
> 35: TEX temp[11].xyz, input[2].xy__, 2D[5];
> 36: MUL_SAT temp[5].x, temp[5].x___, const[7].z___;
> 37: MAD temp[1].xyz, temp[11].xyz_, -temp[5].xxx_, temp[11].xyz_;
> 38: MAX temp[1].xyz, temp[1].xyz_, none.000_;
> 39: ADD temp[10].xyz, temp[1].xyz_, temp[10].xyz_;
> 40: ADD output[0].xyz, temp[3].xyz_, temp[10].xyz_;
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: TEX temp[0], input[2].xy__, 2D[0];
>  1: ADD temp[1].w, temp[0].___w, -const[6].___x;
>  2: MOV temp[12], temp[1].wwww;
>  3: KIL temp[12];
>  4: TEX temp[2].yw, input[2].xy__, 2D[1];
>  5: MOV temp[2].x, temp[2].w___;
>  6: MAD temp[2].xy, temp[2].xy__, const[7].xx__, none.-1-1__;
>  7: MOV temp[2].z, none.__0_;
>  8: DP3 temp[3].x, temp[2].xyz_, temp[2].xyz_;
>  9: ADD temp[3].x, none.1___, -temp[3].x___;
> 10: RSQ temp[3].x, |temp[3].x___|;
> 11: RCP temp[2].z, temp[3].x___;
> 12: DP3 temp[4].x, input[3].xyz_, temp[2].xyz_;
> 13: DP3 temp[4].y, input[4].xyz_, temp[2].xyz_;
> 14: DP3 temp[4].z, input[5].xyz_, temp[2].xyz_;
> 15: TEX temp[5].xyz, temp[4].xyz_, CUBE[2];
> 16: MUL temp[5].xyz, temp[5].xyz_, const[3].xyz_;
> 17: MUL temp[5].xyz, temp[5].xyz_, const[7].xxx_;
> 18: MUL temp[6].xyz, temp[0].xyz_, const[4].xyz_;
> 19: MUL temp[6].xyz, temp[5].xyz_, temp[6].xyz_;
> 20: DP3 temp[3].x, input[1].xyz_, temp[4].xyz_;
> 21: ADD temp[3].x, temp[3].x___, temp[3].x___;
> 22: MAD temp[7].xyz, temp[4].xyz_, temp[3].xxx_, -input[1].xyz_;
> 23: TEX temp[8].xyz, temp[7].xyz_, CUBE[3];
> 24: TEX temp[9].xyz, input[2].xy__, 2D[4];
> 25: ADD temp[1].xyz, temp[9].xyz_, temp[9].xyz_;
> 26: MAD temp[6].xyz, temp[1].xyz_, temp[8].xyz_, temp[6].xyz_;
> 27: MUL temp[3].xyz, temp[6].xyz_, input[0].xyz_;
> 28: MOV output[0].w, temp[0].___w;
> 29: DP3 temp[1].xyz, input[1].xyz_, temp[4].xyz_;
> 30: ADD temp[1].xyz, none.111_, -temp[1].xyz_;
> 31: MUL temp[1].xyz, temp[1].xyz_, temp[1].xyz_;
> 32: MUL temp[1].xyz, temp[1].xyz_, temp[1].xyz_;
> 33: MUL temp[1].xyz, temp[1].xyz_, const[5].xyz_;
> 34: MUL temp[10].xyz, temp[1].xyz_, temp[5].xyz_;
> 35: MAD temp[10].xyz, temp[1].xyz_, const[6].www_, temp[10].xyz_;
> 36: TEX temp[11].xyz, input[2].xy__, 2D[5];
> 37: MUL_SAT temp[5].x, temp[5].x___, const[7].z___;
> 38: MAD temp[1].xyz, temp[11].xyz_, -temp[5].xxx_, temp[11].xyz_;
> 39: MAX temp[1].xyz, temp[1].xyz_, none.000_;
> 40: ADD temp[10].xyz, temp[1].xyz_, temp[10].xyz_;
> 41: ADD output[0].xyz, temp[3].xyz_, temp[10].xyz_;
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: TEX temp[0], input[2].xy__, 2D[0];
>  1: src0.xyz = const[6], src0.w = temp[0]
>     MAD temp[1].w, src0.w, src0.1, -src0.x
>  2: src0.w = temp[1]
>     MAD temp[12].xyz, src0.www, src0.111, src0.000
>     MAD temp[12].w, src0.w, src0.1, src0.0
>  3: KIL temp[12];
>  4: TEX temp[2].yw, input[2].xy__, 2D[1];
>  5: src0.w = temp[2]
>     MAD temp[2].x, src0.w__, src0.111, src0.000
>  6: src0.xyz = temp[2], src1.xyz = const[7]
>     MAD temp[2].xy, src0.xy_, src1.xx_, -src0.11_
>  7: 
>     MAD temp[2].z, src0.__0, src0.111, src0.000
>  8: src0.xyz = temp[2]
>     DP3 temp[3].x, src0.xyz, src0.xyz
>  9: src0.xyz = temp[3]
>     MAD temp[3].x, src0.1__, src0.111, -src0.x__
> 10: src0.xyz = temp[3]
>     REPL_ALPHA temp[3].x
>     RSQ, |src0.x|
> 11: src0.xyz = temp[3]
>     REPL_ALPHA temp[2].z
>     RCP, src0.x
> 12: src0.xyz = input[3], src1.xyz = temp[2]
>     DP3 temp[4].x, src0.xyz, src1.xyz
> 13: src0.xyz = input[4], src1.xyz = temp[2]
>     DP3 temp[4].y, src0.xyz, src1.xyz
> 14: src0.xyz = input[5], src1.xyz = temp[2]
>     DP3 temp[4].z, src0.xyz, src1.xyz
> 15: TEX temp[5].xyz, temp[4].xyz_, CUBE[2];
> 16: src0.xyz = temp[5], src1.xyz = const[3]
>     MAD temp[5].xyz, src0.xyz, src1.xyz, src0.000
> 17: src0.xyz = temp[5], src1.xyz = const[7]
>     MAD temp[5].xyz, src0.xyz, src1.xxx, src0.000
> 18: src0.xyz = temp[0], src1.xyz = const[4]
>     MAD temp[6].xyz, src0.xyz, src1.xyz, src0.000
> 19: src0.xyz = temp[5], src1.xyz = temp[6]
>     MAD temp[6].xyz, src0.xyz, src1.xyz, src0.000
> 20: src0.xyz = input[1], src1.xyz = temp[4]
>     DP3 temp[3].x, src0.xyz, src1.xyz
> 21: src0.xyz = temp[3]
>     MAD temp[3].x, src0.x__, src0.111, src0.x__
> 22: src0.xyz = temp[4], src1.xyz = temp[3], src2.xyz = input[1]
>     MAD temp[7].xyz, src0.xyz, src1.xxx, -src2.xyz
> 23: TEX temp[8].xyz, temp[7].xyz_, CUBE[3];
> 24: TEX temp[9].xyz, input[2].xy__, 2D[4];
> 25: src0.xyz = temp[9]
>     MAD temp[1].xyz, src0.xyz, src0.111, src0.xyz
> 26: src0.xyz = temp[1], src1.xyz = temp[8], src2.xyz = temp[6]
>     MAD temp[6].xyz, src0.xyz, src1.xyz, src2.xyz
> 27: src0.xyz = temp[6], src1.xyz = input[0]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src0.000
> 28: src0.w = temp[0]
>     MAD color[0].w, src0.w, src0.1, src0.0
> 29: src0.xyz = input[1], src1.xyz = temp[4]
>     DP3 temp[1].xyz, src0.xyz, src1.xyz
> 30: src0.xyz = temp[1]
>     MAD temp[1].xyz, src0.111, src0.111, -src0.xyz
> 31: src0.xyz = temp[1]
>     MAD temp[1].xyz, src0.xyz, src0.xyz, src0.000
> 32: src0.xyz = temp[1]
>     MAD temp[1].xyz, src0.xyz, src0.xyz, src0.000
> 33: src0.xyz = temp[1], src1.xyz = const[5]
>     MAD temp[1].xyz, src0.xyz, src1.xyz, src0.000
> 34: src0.xyz = temp[1], src1.xyz = temp[5]
>     MAD temp[10].xyz, src0.xyz, src1.xyz, src0.000
> 35: src0.xyz = temp[1], src0.w = const[6], src1.xyz = temp[10]
>     MAD temp[10].xyz, src0.xyz, src0.www, src1.xyz
> 36: TEX temp[11].xyz, input[2].xy__, 2D[5];
> 37: src0.xyz = temp[5], src1.xyz = const[7]
>     MAD_SAT temp[5].x, src0.x__, src1.z__, src0.000
> 38: src0.xyz = temp[11], src1.xyz = temp[5]
>     MAD temp[1].xyz, src0.xyz, -src1.xxx, src0.xyz
> 39: src0.xyz = temp[1]
>     MAX temp[1].xyz, src0.xyz, src0.000
> 40: src0.xyz = temp[1], src1.xyz = temp[10]
>     MAD temp[10].xyz, src0.xyz, src0.111, src1.xyz
> 41: src0.xyz = temp[3], src1.xyz = temp[10]
>     MAD color[0].xyz, src0.xyz, src0.111, src1.xyz
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[11].xyz, input[2].xy__, 2D[5];
>  2: TEX temp[9].xyz, input[2].xy__, 2D[4];
>  3: TEX temp[2].yw, input[2].xy__, 2D[1];
>  4: TEX temp[0], input[2].xy__, 2D[0];
>  5: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[4], src2.xyz = const[6]
>     MAD temp[6].xyz, src0.xyz, src1.xyz, src0.000
>     MAD temp[1].w, src0.w, src0.1, -src2.x
>  6: src0.w = temp[1]
>     MAD temp[12].xyz, src0.www, src0.111, src0.000
>     MAD temp[12].w, src0.w, src0.1, src0.0
>  7: src0.w = temp[2], src1.w = temp[0]
>     MAD temp[2].x, src0.w__, src0.111, src0.000
>     MAD color[0].w, src1.w, src0.1, src0.0
>  8: src0.xyz = temp[2], src1.xyz = const[7]
>     MAD temp[2].xy, src0.xy_, src1.xx_, -src0.11_
>  9: src0.xyz = temp[9]
>     MAD temp[1].xyz, src0.xyz, src0.111, src0.xyz
> 10: 
>     MAD temp[2].z, src0.__0, src0.111, src0.000
> 11: src0.xyz = temp[2]
>     DP3 temp[3].x, src0.xyz, src0.xyz
> 12: src0.xyz = temp[3]
>     MAD temp[3].x, src0.1__, src0.111, -src0.x__
> 13: src0.xyz = temp[3]
>     REPL_ALPHA temp[3].x
>     RSQ, |src0.x|
> 14: src0.xyz = temp[3]
>     REPL_ALPHA temp[2].z
>     RCP, src0.x
> 15: src0.xyz = input[3], src1.xyz = temp[2]
>     DP3 temp[4].x, src0.xyz, src1.xyz
> 16: src0.xyz = input[4], src1.xyz = temp[2]
>     DP3 temp[4].y, src0.xyz, src1.xyz
> 17: src0.xyz = input[5], src1.xyz = temp[2]
>     DP3 temp[4].z, src0.xyz, src1.xyz
> 18: src0.xyz = input[1], src1.xyz = temp[4]
>     DP3 temp[3].x, src0.xyz, src1.xyz
> 19: src0.xyz = temp[3]
>     MAD temp[3].x, src0.x__, src0.111, src0.x__
> 20: src0.xyz = temp[4], src1.xyz = temp[3], src2.xyz = input[1]
>     MAD temp[7].xyz, src0.xyz, src1.xxx, -src2.xyz
> 21: BEGIN_TEX;
> 22: TEX temp[8].xyz, temp[7].xyz_, CUBE[3];
> 23: TEX temp[5].xyz, temp[4].xyz_, CUBE[2];
> 24: KIL temp[12];
> 25: src0.xyz = temp[5], src1.xyz = const[3]
>     MAD temp[5].xyz, src0.xyz, src1.xyz, src0.000
> 26: src0.xyz = temp[5], src1.xyz = const[7]
>     MAD temp[5].xyz, src0.xyz, src1.xxx, src0.000
> 27: src0.xyz = temp[5], src1.xyz = temp[6]
>     MAD temp[6].xyz, src0.xyz, src1.xyz, src0.000
> 28: src0.xyz = temp[1], src1.xyz = temp[8], src2.xyz = temp[6]
>     MAD temp[6].xyz, src0.xyz, src1.xyz, src2.xyz
> 29: src0.xyz = temp[6], src1.xyz = input[0]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src0.000
> 30: src0.xyz = input[1], src1.xyz = temp[4]
>     DP3 temp[1].xyz, src0.xyz, src1.xyz
> 31: src0.xyz = temp[1]
>     MAD temp[1].xyz, src0.111, src0.111, -src0.xyz
> 32: src0.xyz = temp[1]
>     MAD temp[1].xyz, src0.xyz, src0.xyz, src0.000
> 33: src0.xyz = temp[1]
>     MAD temp[1].xyz, src0.xyz, src0.xyz, src0.000
> 34: src0.xyz = temp[1], src1.xyz = const[5]
>     MAD temp[1].xyz, src0.xyz, src1.xyz, src0.000
> 35: src0.xyz = temp[1], src1.xyz = temp[5]
>     MAD temp[10].xyz, src0.xyz, src1.xyz, src0.000
> 36: src0.xyz = temp[1], src0.w = const[6], src1.xyz = temp[10]
>     MAD temp[10].xyz, src0.xyz, src0.www, src1.xyz
> 37: src0.xyz = temp[5], src1.xyz = const[7]
>     MAD_SAT temp[5].x, src0.x__, src1.z__, src0.000
> 38: src0.xyz = temp[11], src1.xyz = temp[5]
>     MAD temp[1].xyz, src0.xyz, -src1.xxx, src0.xyz
> 39: src0.xyz = temp[1]
>     MAX temp[1].xyz, src0.xyz, src0.000
> 40: src0.xyz = temp[1], src1.xyz = temp[10]
>     MAD temp[10].xyz, src0.xyz, src0.111, src1.xyz
> 41: src0.xyz = temp[3], src1.xyz = temp[10]
>     MAD color[0].xyz, src0.xyz, src0.111, src1.xyz
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[10].xyz, temp[2].xy__, 2D[5];
>  2: TEX temp[9].xyz, temp[2].xy__, 2D[4];
>  3: TEX temp[7].yw, temp[2].xy__, 2D[1];
>  4: TEX temp[2], temp[2].xy__, 2D[0];
>  5: src0.xyz = temp[2], src0.w = temp[2], src1.xyz = const[4], src2.xyz = const[6]
>     MAD temp[8].xyz, src0.xyz, src1.xyz, src0.000
>     MAD temp[6].w, src0.w, src0.1, -src2.x
>  6: src0.w = temp[6]
>     MAD temp[11].xyz, src0.www, src0.111, src0.000
>     MAD temp[11].w, src0.w, src0.1, src0.0
>  7: src0.w = temp[7], src1.w = temp[2]
>     MAD temp[7].x, src0.w__, src0.111, src0.000
>     MAD color[0].w, src1.w, src0.1, src0.0
>  8: src0.xyz = temp[7], src1.xyz = const[7]
>     MAD temp[7].xy, src0.xy_, src1.xx_, -src0.11_
>  9: src0.xyz = temp[9]
>     MAD temp[6].xyz, src0.xyz, src0.111, src0.xyz
> 10: 
>     MAD temp[7].z, src0.__0, src0.111, src0.000
> 11: src0.xyz = temp[7]
>     DP3 temp[2].x, src0.xyz, src0.xyz
> 12: src0.xyz = temp[2]
>     MAD temp[2].x, src0.1__, src0.111, -src0.x__
> 13: src0.xyz = temp[2]
>     REPL_ALPHA temp[2].x
>     RSQ, |src0.x|
> 14: src0.xyz = temp[2]
>     REPL_ALPHA temp[7].z
>     RCP, src0.x
> 15: src0.xyz = temp[3], src1.xyz = temp[7]
>     DP3 temp[3].x, src0.xyz, src1.xyz
> 16: src0.xyz = temp[4], src1.xyz = temp[7]
>     DP3 temp[3].y, src0.xyz, src1.xyz
> 17: src0.xyz = temp[5], src1.xyz = temp[7]
>     DP3 temp[3].z, src0.xyz, src1.xyz
> 18: src0.xyz = temp[1], src1.xyz = temp[3]
>     DP3 temp[2].x, src0.xyz, src1.xyz
> 19: src0.xyz = temp[2]
>     MAD temp[2].x, src0.x__, src0.111, src0.x__
> 20: src0.xyz = temp[3], src1.xyz = temp[2], src2.xyz = temp[1]
>     MAD temp[4].xyz, src0.xyz, src1.xxx, -src2.xyz
> 21: BEGIN_TEX;
> 22: TEX temp[5].xyz, temp[4].xyz_, CUBE[3];
> 23: TEX temp[4].xyz, temp[3].xyz_, CUBE[2];
> 24: KIL temp[11];
> 25: src0.xyz = temp[4], src1.xyz = const[3]
>     MAD temp[4].xyz, src0.xyz, src1.xyz, src0.000
> 26: src0.xyz = temp[4], src1.xyz = const[7]
>     MAD temp[4].xyz, src0.xyz, src1.xxx, src0.000
> 27: src0.xyz = temp[4], src1.xyz = temp[8]
>     MAD temp[8].xyz, src0.xyz, src1.xyz, src0.000
> 28: src0.xyz = temp[6], src1.xyz = temp[5], src2.xyz = temp[8]
>     MAD temp[8].xyz, src0.xyz, src1.xyz, src2.xyz
> 29: src0.xyz = temp[8], src1.xyz = temp[0]
>     MAD temp[2].xyz, src0.xyz, src1.xyz, src0.000
> 30: src0.xyz = temp[1], src1.xyz = temp[3]
>     DP3 temp[6].xyz, src0.xyz, src1.xyz
> 31: src0.xyz = temp[6]
>     MAD temp[6].xyz, src0.111, src0.111, -src0.xyz
> 32: src0.xyz = temp[6]
>     MAD temp[6].xyz, src0.xyz, src0.xyz, src0.000
> 33: src0.xyz = temp[6]
>     MAD temp[6].xyz, src0.xyz, src0.xyz, src0.000
> 34: src0.xyz = temp[6], src1.xyz = const[5]
>     MAD temp[6].xyz, src0.xyz, src1.xyz, src0.000
> 35: src0.xyz = temp[6], src1.xyz = temp[4]
>     MAD temp[0].xyz, src0.xyz, src1.xyz, src0.000
> 36: src0.xyz = temp[6], src0.w = const[6], src1.xyz = temp[0]
>     MAD temp[0].xyz, src0.xyz, src0.www, src1.xyz
> 37: src0.xyz = temp[4], src1.xyz = const[7]
>     MAD_SAT temp[4].x, src0.x__, src1.z__, src0.000
> 38: src0.xyz = temp[10], src1.xyz = temp[4]
>     MAD temp[6].xyz, src0.xyz, -src1.xxx, src0.xyz
> 39: src0.xyz = temp[6]
>     MAX temp[6].xyz, src0.xyz, src0.000
> 40: src0.xyz = temp[6], src1.xyz = temp[0]
>     MAD temp[0].xyz, src0.xyz, src0.111, src1.xyz
> 41: src0.xyz = temp[2], src1.xyz = temp[0]
>     MAD color[0].xyz, src0.xyz, src0.111, src1.xyz
>pc=6*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 15, tex_end: 3  (code_addr: 004603c0)
>  TEX:
>    TEX t10, t2, texture[5]   (0000aa82)
>    TEX t9, t2, texture[4]   (0000a242)
>    TEX t7, t2, texture[1]   (000089c2)
>    TEX t2, t2, texture[0]   (00008082)
>  0: xyz:  t2  c4  c6 -> t8.xyz               (03a26902)
>       w:  t2  t0  t0 -> t6.w                 (00980002)
>     xyz:   t2.xyz   c4.xyz      0.0    op: 00050200
>       w:     t2.w      1.0    -c6.x    op: 00098889
>  1: xyz:  t0  t0  t0 -> t11.xyz              (03ac0000)
>       w:  t6  t0  t0 -> t11.w                (00ac0006)
>     xyz:   t6.www      1.0      0.0    op: 00050a8c
>       w:     t6.w      1.0      0.0    op: 00040889
>  2: xyz:  t0  t0  t0 -> t7.x                 (009c0000)
>       w:  t7  t2  t0 -> o0.w                 (01000087)
>     xyz:   t7.www      1.0      0.0    op: 00050a8c
>       w:     t2.w      1.0      0.0    op: 0004088a
>  3: xyz:  t7  c7  t0 -> t7.xy                (019c09c7)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t7.xyz   c7.xxx     -1.0    op: 000d4280
>       w:     t7.x     t7.x     t7.x    op: 00000000
>  4: xyz:  t9  t0  t0 -> t6.xyz               (03980009)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t9.xyz      1.0   t9.xyz    op: 00000a80
>       w:     t9.x     t9.x     t9.x    op: 00000000
>  5: xyz:  t0  t0  t0 -> t7.z                 (021c0000)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:      0.0      1.0      0.0    op: 00050a94
>       w:     t0.x     t0.x     t0.x    op: 00000000
>  6: xyz:  t7  t0  t0 -> t2.x                 (00880007)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t7.xyz   t7.xyz   t7.xxx    op: 00804000
>       w:     t7.x     t7.x     t7.x    op: 00000000
>  7: xyz:  t2  t0  t0 -> t2.x                 (00880002)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:      1.0      1.0  -t2.xyz    op: 00080a95
>       w:     t2.x     t2.x     t2.x    op: 00000000
>  8: xyz:  t2  t0  t0 -> t2.x                 (00880002)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xxx   t2.xxx   t2.xxx    op: 05004081
>       w:   |t2.x|     t2.x     t2.x    op: 05800040
>  9: xyz:  t2  t0  t0 -> t7.z                 (021c0002)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xxx   t2.xxx   t2.xxx    op: 05004081
>       w:     t2.x     t2.x     t2.x    op: 05000000
> 10: xyz:  t3  t7  t0 -> t3.x                 (008c01c3)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t3.xyz   t7.xyz   t3.xxx    op: 00804200
>       w:     t3.x     t3.x     t3.x    op: 00000000
> 11: xyz:  t4  t7  t0 -> t3.y                 (010c01c4)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t4.xyz   t7.xyz   t4.xxx    op: 00804200
>       w:     t4.x     t4.x     t4.x    op: 00000000
> 12: xyz:  t5  t7  t0 -> t3.z                 (020c01c5)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t5.xyz   t7.xyz   t5.xxx    op: 00804200
>       w:     t5.x     t5.x     t5.x    op: 00000000
> 13: xyz:  t1  t3  t0 -> t2.x                 (008800c1)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   t3.xyz   t1.xxx    op: 00804200
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 14: xyz:  t2  t0  t0 -> t2.x                 (00880002)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz      1.0   t2.xyz    op: 00000a80
>       w:     t2.x     t2.x     t2.x    op: 00000000
> 15: xyz:  t3  t2  t1 -> t4.xyz               (03901083)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t3.xyz   t2.xxx  -t1.xyz    op: 000a0280
>       w:     t3.x     t3.x     t3.x    op: 00000000
>NODE 1: alu_offset: 16, tex_offset: 4, alu_end: 16, tex_end: 2  (code_addr: 00444410)
>  TEX:
>    TEX t5, t4, texture[3]   (00009944)
>    TEX t4, t3, texture[2]   (00009103)
>    KIL t0, t11, texture[0]   (0001000b)
> 16: xyz:  t4  c3  t0 -> t4.xyz               (039008c4)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t4.xyz   c3.xyz      0.0    op: 00050200
>       w:     t4.x     t4.x     t4.x    op: 00000000
> 17: xyz:  t4  c7  t0 -> t4.xyz               (039009c4)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t4.xyz   c7.xxx      0.0    op: 00050280
>       w:     t4.x     t4.x     t4.x    op: 00000000
> 18: xyz:  t4  t8  t0 -> t8.xyz               (03a00204)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t4.xyz   t8.xyz      0.0    op: 00050200
>       w:     t4.x     t4.x     t4.x    op: 00000000
> 19: xyz:  t6  t5  t8 -> t8.xyz               (03a08146)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t6.xyz   t5.xyz   t8.xyz    op: 00020200
>       w:     t6.x     t6.x     t6.x    op: 00000000
> 20: xyz:  t8  t0  t0 -> t2.xyz               (03880008)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t8.xyz   t0.xyz      0.0    op: 00050200
>       w:     t8.x     t8.x     t8.x    op: 00000000
> 21: xyz:  t1  t3  t0 -> t6.xyz               (039800c1)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   t3.xyz   t1.xxx    op: 00804200
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 22: xyz:  t6  t0  t0 -> t6.xyz               (03980006)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:      1.0      1.0  -t6.xyz    op: 00080a95
>       w:     t6.x     t6.x     t6.x    op: 00000000
> 23: xyz:  t6  t0  t0 -> t6.xyz               (03980006)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t6.xyz   t6.xyz      0.0    op: 00050000
>       w:     t6.x     t6.x     t6.x    op: 00000000
> 24: xyz:  t6  t0  t0 -> t6.xyz               (03980006)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t6.xyz   t6.xyz      0.0    op: 00050000
>       w:     t6.x     t6.x     t6.x    op: 00000000
> 25: xyz:  t6  c5  t0 -> t6.xyz               (03980946)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t6.xyz   c5.xyz      0.0    op: 00050200
>       w:     t6.x     t6.x     t6.x    op: 00000000
> 26: xyz:  t6  t4  t0 -> t0.xyz               (03800106)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t6.xyz   t4.xyz      0.0    op: 00050200
>       w:     t6.x     t6.x     t6.x    op: 00000000
> 27: xyz:  t6  t0  t0 -> t0.xyz               (03800006)
>       w:  c6  t0  t0 ->                      (00000026)
>     xyz:   t6.xyz   c6.www   t0.xyz    op: 00010600
>       w:     t6.x     t6.x     t6.x    op: 00000000
> 28: xyz:  t4  c7  t0 -> t4.x                 (009009c4)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t4.xyz   c7.zzz      0.0    op: 40050380
>       w:     t4.x     t4.x     t4.x    op: 00000000
> 29: xyz: t10  t4  t0 -> t6.xyz               (0398010a)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:  t10.xyz  -t4.xxx  t10.xyz    op: 00001280
>       w:    t10.x    t10.x    t10.x    op: 00000000
> 30: xyz:  t6  t0  t0 -> t6.xyz               (03980006)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t6.xyz      0.0   t6.xxx    op: 02804a00
>       w:     t6.x     t6.x     t6.x    op: 00000000
> 31: xyz:  t6  t0  t0 -> t0.xyz               (03800006)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t6.xyz      1.0   t0.xyz    op: 00010a80
>       w:     t6.x     t6.x     t6.x    op: 00000000
> 32: xyz:  t2  t0  t0 -> o0.xyz               (1c000002)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz      1.0   t0.xyz    op: 00010a80
>       w:     t2.x     t2.x     t2.x    op: 00000000
>r300: Initial fragment program
>FRAG
>DCL IN[0], COLOR, LINEAR
>DCL IN[1], GENERIC[2], PERSPECTIVE
>DCL IN[2], GENERIC[4], PERSPECTIVE
>DCL IN[3], GENERIC[5], PERSPECTIVE
>DCL IN[4], GENERIC[6], PERSPECTIVE
>DCL IN[5], GENERIC[7], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL SAMP[1]
>DCL SAMP[2]
>DCL SAMP[3]
>DCL SAMP[4]
>DCL SAMP[5]
>DCL CONST[3..5]
>DCL TEMP[0..12]
>IMM FLT32 {    0.0010,     0.0000,     1.0000,     0.1000}
>IMM FLT32 {    2.0000,    -1.0000,     8.0000,     0.0000}
>  0: TEX TEMP[0], IN[2], SAMP[0], 2D
>  1: SUB TEMP[1].w, TEMP[0].wwww, IMM[0].xxxx
>  2: TEX TEMP[2], IN[2], SAMP[1], 2D
>  3: MOV TEMP[2].x, TEMP[2].wwww
>  4: MAD TEMP[2].xy, TEMP[2], IMM[1].xxxx, IMM[1].yyyy
>  5: MOV TEMP[2].z, IMM[0].yyyy
>  6: DP3 TEMP[3].x, TEMP[2], TEMP[2]
>  7: ADD TEMP[3].x, IMM[0].zzzz, -TEMP[3].xxxx
>  8: RSQ TEMP[3].x, TEMP[3].xxxx
>  9: RCP TEMP[2].z, TEMP[3].xxxx
> 10: DP3 TEMP[4].x, IN[3], TEMP[2]
> 11: DP3 TEMP[4].y, IN[4], TEMP[2]
> 12: DP3 TEMP[4].z, IN[5], TEMP[2]
> 13: TEX TEMP[5], TEMP[4], SAMP[2], CUBE
> 14: MUL TEMP[5], TEMP[5], CONST[3]
> 15: MUL TEMP[5], TEMP[5], IMM[1].xxxx
> 16: MUL TEMP[6], TEMP[0], CONST[4]
> 17: MUL TEMP[6], TEMP[5], TEMP[6]
> 18: DP3 TEMP[3], IN[1], TEMP[4]
> 19: ADD TEMP[3], TEMP[3], TEMP[3]
> 20: MAD TEMP[7], TEMP[4], TEMP[3].xxxx, -IN[1]
> 21: TEX TEMP[8], TEMP[7], SAMP[3], CUBE
> 22: TEX TEMP[9], IN[2], SAMP[4], 2D
> 23: ADD TEMP[1], TEMP[9], TEMP[9]
> 24: MAD TEMP[6], TEMP[1], TEMP[8], TEMP[6]
> 25: MUL TEMP[3], TEMP[6], IN[0]
> 26: DP3 TEMP[1], IN[1], TEMP[4]
> 27: ADD TEMP[1], IMM[0].zzzz, -TEMP[1]
> 28: MUL TEMP[1], TEMP[1], TEMP[1]
> 29: MUL TEMP[1], TEMP[1], TEMP[1]
> 30: MUL TEMP[10], TEMP[1], CONST[5]
> 31: MUL TEMP[11], TEMP[10], TEMP[5]
> 32: MAD TEMP[11], TEMP[10], IMM[0].wwww, TEMP[11]
> 33: TEX TEMP[12], IN[2], SAMP[5], 2D
> 34: MUL_SAT TEMP[5], TEMP[5], IMM[1].zzzz
> 35: MAD TEMP[1], TEMP[12], -TEMP[5].xxxx, TEMP[12]
> 36: ADD OUT[0].xyz, TEMP[3], TEMP[11]
> 37: SUB TEMP[1], IMM[0].zzzz, TEMP[10]
> 38: MUL OUT[0].w, TEMP[0].wwww, TEMP[1]
> 39: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: TEX temp[0], input[2], 2D[0];
>  1: SUB temp[1].w, temp[0].wwww, const[6].xxxx;
>  2: TEX temp[2], input[2], 2D[1];
>  3: MOV temp[2].x, temp[2].wwww;
>  4: MAD temp[2].xy, temp[2], const[7].xxxx, const[7].yyyy;
>  5: MOV temp[2].z, const[6].yyyy;
>  6: DP3 temp[3].x, temp[2], temp[2];
>  7: ADD temp[3].x, const[6].zzzz, -temp[3].xxxx;
>  8: RSQ temp[3].x, temp[3].xxxx;
>  9: RCP temp[2].z, temp[3].xxxx;
> 10: DP3 temp[4].x, input[3], temp[2];
> 11: DP3 temp[4].y, input[4], temp[2];
> 12: DP3 temp[4].z, input[5], temp[2];
> 13: TEX temp[5], temp[4], CUBE[2];
> 14: MUL temp[5], temp[5], const[3];
> 15: MUL temp[5], temp[5], const[7].xxxx;
> 16: MUL temp[6], temp[0], const[4];
> 17: MUL temp[6], temp[5], temp[6];
> 18: DP3 temp[3], input[1], temp[4];
> 19: ADD temp[3], temp[3], temp[3];
> 20: MAD temp[7], temp[4], temp[3].xxxx, -input[1];
> 21: TEX temp[8], temp[7], CUBE[3];
> 22: TEX temp[9], input[2], 2D[4];
> 23: ADD temp[1], temp[9], temp[9];
> 24: MAD temp[6], temp[1], temp[8], temp[6];
> 25: MUL temp[3], temp[6], input[0];
> 26: DP3 temp[1], input[1], temp[4];
> 27: ADD temp[1], const[6].zzzz, -temp[1];
> 28: MUL temp[1], temp[1], temp[1];
> 29: MUL temp[1], temp[1], temp[1];
> 30: MUL temp[10], temp[1], const[5];
> 31: MUL temp[11], temp[10], temp[5];
> 32: MAD temp[11], temp[10], const[6].wwww, temp[11];
> 33: TEX temp[12], input[2], 2D[5];
> 34: MUL_SAT temp[5], temp[5], const[7].zzzz;
> 35: MAD temp[1], temp[12], -temp[5].xxxx, temp[12];
> 36: ADD output[0].xyz, temp[3], temp[11];
> 37: SUB temp[1], const[6].zzzz, temp[10];
> 38: MUL output[0].w, temp[0].wwww, temp[1];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: TEX temp[0], input[2], 2D[0];
>  1: SUB temp[1].w, temp[0].wwww, const[6].xxxx;
>  2: TEX temp[2], input[2], 2D[1];
>  3: MOV temp[2].x, temp[2].wwww;
>  4: MAD temp[2].xy, temp[2], const[7].xxxx, const[7].yyyy;
>  5: MOV temp[2].z, const[6].yyyy;
>  6: DP3 temp[3].x, temp[2], temp[2];
>  7: ADD temp[3].x, const[6].zzzz, -temp[3].xxxx;
>  8: RSQ temp[3].x, temp[3].xxxx;
>  9: RCP temp[2].z, temp[3].xxxx;
> 10: DP3 temp[4].x, input[3], temp[2];
> 11: DP3 temp[4].y, input[4], temp[2];
> 12: DP3 temp[4].z, input[5], temp[2];
> 13: TEX temp[5], temp[4], CUBE[2];
> 14: MUL temp[5], temp[5], const[3];
> 15: MUL temp[5], temp[5], const[7].xxxx;
> 16: MUL temp[6], temp[0], const[4];
> 17: MUL temp[6], temp[5], temp[6];
> 18: DP3 temp[3], input[1], temp[4];
> 19: ADD temp[3], temp[3], temp[3];
> 20: MAD temp[7], temp[4], temp[3].xxxx, -input[1];
> 21: TEX temp[8], temp[7], CUBE[3];
> 22: TEX temp[9], input[2], 2D[4];
> 23: ADD temp[1], temp[9], temp[9];
> 24: MAD temp[6], temp[1], temp[8], temp[6];
> 25: MUL temp[3], temp[6], input[0];
> 26: DP3 temp[1], input[1], temp[4];
> 27: ADD temp[1], const[6].zzzz, -temp[1];
> 28: MUL temp[1], temp[1], temp[1];
> 29: MUL temp[1], temp[1], temp[1];
> 30: MUL temp[10], temp[1], const[5];
> 31: MUL temp[11], temp[10], temp[5];
> 32: MAD temp[11], temp[10], const[6].wwww, temp[11];
> 33: TEX temp[12], input[2], 2D[5];
> 34: MUL_SAT temp[5], temp[5], const[7].zzzz;
> 35: MAD temp[1], temp[12], -temp[5].xxxx, temp[12];
> 36: ADD output[0].xyz, temp[3], temp[11];
> 37: SUB temp[1], const[6].zzzz, temp[10];
> 38: MUL output[0].w, temp[0].wwww, temp[1];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: TEX temp[0], input[2], 2D[0];
>  1: SUB temp[1].w, temp[0].wwww, const[6].xxxx;
>  2: TEX temp[2], input[2], 2D[1];
>  3: MOV temp[2].x, temp[2].wwww;
>  4: MAD temp[2].xy, temp[2], const[7].xxxx, const[7].yyyy;
>  5: MOV temp[2].z, const[6].yyyy;
>  6: DP3 temp[3].x, temp[2], temp[2];
>  7: ADD temp[3].x, const[6].zzzz, -temp[3].xxxx;
>  8: RSQ temp[3].x, temp[3].xxxx;
>  9: RCP temp[2].z, temp[3].xxxx;
> 10: DP3 temp[4].x, input[3], temp[2];
> 11: DP3 temp[4].y, input[4], temp[2];
> 12: DP3 temp[4].z, input[5], temp[2];
> 13: TEX temp[5], temp[4], CUBE[2];
> 14: MUL temp[5], temp[5], const[3];
> 15: MUL temp[5], temp[5], const[7].xxxx;
> 16: MUL temp[6], temp[0], const[4];
> 17: MUL temp[6], temp[5], temp[6];
> 18: DP3 temp[3], input[1], temp[4];
> 19: ADD temp[3], temp[3], temp[3];
> 20: MAD temp[7], temp[4], temp[3].xxxx, -input[1];
> 21: TEX temp[8], temp[7], CUBE[3];
> 22: TEX temp[9], input[2], 2D[4];
> 23: ADD temp[1], temp[9], temp[9];
> 24: MAD temp[6], temp[1], temp[8], temp[6];
> 25: MUL temp[3], temp[6], input[0];
> 26: DP3 temp[1], input[1], temp[4];
> 27: ADD temp[1], const[6].zzzz, -temp[1];
> 28: MUL temp[1], temp[1], temp[1];
> 29: MUL temp[1], temp[1], temp[1];
> 30: MUL temp[10], temp[1], const[5];
> 31: MUL temp[11], temp[10], temp[5];
> 32: MAD temp[11], temp[10], const[6].wwww, temp[11];
> 33: TEX temp[12], input[2], 2D[5];
> 34: MUL_SAT temp[5], temp[5], const[7].zzzz;
> 35: MAD temp[1], temp[12], -temp[5].xxxx, temp[12];
> 36: ADD output[0].xyz, temp[3], temp[11];
> 37: SUB temp[1], const[6].zzzz, temp[10];
> 38: MUL output[0].w, temp[0].wwww, temp[1];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: TEX temp[0], input[2], 2D[0];
>  1: ADD temp[1].w, temp[0].wwww, -const[6].xxxx;
>  2: TEX temp[2], input[2], 2D[1];
>  3: MOV temp[2].x, temp[2].wwww;
>  4: MAD temp[2].xy, temp[2], const[7].xxxx, const[7].yyyy;
>  5: MOV temp[2].z, const[6].yyyy;
>  6: DP3 temp[3].x, temp[2], temp[2];
>  7: ADD temp[3].x, const[6].zzzz, -temp[3].xxxx;
>  8: RSQ temp[3].x, |temp[3].xxxx|;
>  9: RCP temp[2].z, temp[3].xxxx;
> 10: DP3 temp[4].x, input[3], temp[2];
> 11: DP3 temp[4].y, input[4], temp[2];
> 12: DP3 temp[4].z, input[5], temp[2];
> 13: TEX temp[5], temp[4], CUBE[2];
> 14: MUL temp[5], temp[5], const[3];
> 15: MUL temp[5], temp[5], const[7].xxxx;
> 16: MUL temp[6], temp[0], const[4];
> 17: MUL temp[6], temp[5], temp[6];
> 18: DP3 temp[3], input[1], temp[4];
> 19: ADD temp[3], temp[3], temp[3];
> 20: MAD temp[7], temp[4], temp[3].xxxx, -input[1];
> 21: TEX temp[8], temp[7], CUBE[3];
> 22: TEX temp[9], input[2], 2D[4];
> 23: ADD temp[1], temp[9], temp[9];
> 24: MAD temp[6], temp[1], temp[8], temp[6];
> 25: MUL temp[3], temp[6], input[0];
> 26: DP3 temp[1], input[1], temp[4];
> 27: ADD temp[1], const[6].zzzz, -temp[1];
> 28: MUL temp[1], temp[1], temp[1];
> 29: MUL temp[1], temp[1], temp[1];
> 30: MUL temp[10], temp[1], const[5];
> 31: MUL temp[11], temp[10], temp[5];
> 32: MAD temp[11], temp[10], const[6].wwww, temp[11];
> 33: TEX temp[12], input[2], 2D[5];
> 34: MUL_SAT temp[5], temp[5], const[7].zzzz;
> 35: MAD temp[1], temp[12], -temp[5].xxxx, temp[12];
> 36: ADD output[0].xyz, temp[3], temp[11];
> 37: ADD temp[1], const[6].zzzz, -temp[10];
> 38: MUL output[0].w, temp[0].wwww, temp[1];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: TEX temp[0], input[2], 2D[0];
>  1: ADD temp[1].w, temp[0].wwww, -const[6].xxxx;
>  2: TEX temp[2], input[2], 2D[1];
>  3: MOV temp[2].x, temp[2].wwww;
>  4: MAD temp[2].xy, temp[2], const[7].xxxx, const[7].yyyy;
>  5: MOV temp[2].z, const[6].yyyy;
>  6: DP3 temp[3].x, temp[2], temp[2];
>  7: ADD temp[3].x, const[6].zzzz, -temp[3].xxxx;
>  8: RSQ temp[3].x, |temp[3].xxxx|;
>  9: RCP temp[2].z, temp[3].xxxx;
> 10: DP3 temp[4].x, input[3], temp[2];
> 11: DP3 temp[4].y, input[4], temp[2];
> 12: DP3 temp[4].z, input[5], temp[2];
> 13: TEX temp[5], temp[4], CUBE[2];
> 14: MUL temp[5], temp[5], const[3];
> 15: MUL temp[5], temp[5], const[7].xxxx;
> 16: MUL temp[6], temp[0], const[4];
> 17: MUL temp[6], temp[5], temp[6];
> 18: DP3 temp[3], input[1], temp[4];
> 19: ADD temp[3], temp[3], temp[3];
> 20: MAD temp[7], temp[4], temp[3].xxxx, -input[1];
> 21: TEX temp[8], temp[7], CUBE[3];
> 22: TEX temp[9], input[2], 2D[4];
> 23: ADD temp[1], temp[9], temp[9];
> 24: MAD temp[6], temp[1], temp[8], temp[6];
> 25: MUL temp[3], temp[6], input[0];
> 26: DP3 temp[1], input[1], temp[4];
> 27: ADD temp[1], const[6].zzzz, -temp[1];
> 28: MUL temp[1], temp[1], temp[1];
> 29: MUL temp[1], temp[1], temp[1];
> 30: MUL temp[10], temp[1], const[5];
> 31: MUL temp[11], temp[10], temp[5];
> 32: MAD temp[11], temp[10], const[6].wwww, temp[11];
> 33: TEX temp[12], input[2], 2D[5];
> 34: MUL_SAT temp[5], temp[5], const[7].zzzz;
> 35: MAD temp[1], temp[12], -temp[5].xxxx, temp[12];
> 36: ADD output[0].xyz, temp[3], temp[11];
> 37: ADD temp[1], const[6].zzzz, -temp[10];
> 38: MUL output[0].w, temp[0].wwww, temp[1];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: TEX temp[0], input[2].xy__, 2D[0];
>  1: TEX temp[2].yw, input[2].xy__, 2D[1];
>  2: MOV temp[2].x, temp[2].w___;
>  3: MAD temp[2].xy, temp[2].xy__, const[7].xx__, const[7].yy__;
>  4: MOV temp[2].z, const[6].__y_;
>  5: DP3 temp[3].x, temp[2].xyz_, temp[2].xyz_;
>  6: ADD temp[3].x, const[6].z___, -temp[3].x___;
>  7: RSQ temp[3].x, |temp[3].x___|;
>  8: RCP temp[2].z, temp[3].x___;
>  9: DP3 temp[4].x, input[3].xyz_, temp[2].xyz_;
> 10: DP3 temp[4].y, input[4].xyz_, temp[2].xyz_;
> 11: DP3 temp[4].z, input[5].xyz_, temp[2].xyz_;
> 12: TEX temp[5].xyz, temp[4].xyz_, CUBE[2];
> 13: MUL temp[5].xyz, temp[5].xyz_, const[3].xyz_;
> 14: MUL temp[5].xyz, temp[5].xyz_, const[7].xxx_;
> 15: MUL temp[6].xyz, temp[0].xyz_, const[4].xyz_;
> 16: MUL temp[6].xyz, temp[5].xyz_, temp[6].xyz_;
> 17: DP3 temp[3].x, input[1].xyz_, temp[4].xyz_;
> 18: ADD temp[3].x, temp[3].x___, temp[3].x___;
> 19: MAD temp[7].xyz, temp[4].xyz_, temp[3].xxx_, -input[1].xyz_;
> 20: TEX temp[8].xyz, temp[7].xyz_, CUBE[3];
> 21: TEX temp[9].xyz, input[2].xy__, 2D[4];
> 22: ADD temp[1].xyz, temp[9].xyz_, temp[9].xyz_;
> 23: MAD temp[6].xyz, temp[1].xyz_, temp[8].xyz_, temp[6].xyz_;
> 24: MUL temp[3].xyz, temp[6].xyz_, input[0].xyz_;
> 25: DP3 temp[1], input[1].xyz_, temp[4].xyz_;
> 26: ADD temp[1], const[6].zzzz, -temp[1];
> 27: MUL temp[1], temp[1], temp[1];
> 28: MUL temp[1], temp[1], temp[1];
> 29: MUL temp[10], temp[1], const[5];
> 30: MUL temp[11].xyz, temp[10].xyz_, temp[5].xyz_;
> 31: MAD temp[11].xyz, temp[10].xyz_, const[6].www_, temp[11].xyz_;
> 32: ADD output[0].xyz, temp[3].xyz_, temp[11].xyz_;
> 33: ADD temp[1].w, const[6].___z, -temp[10].___w;
> 34: MUL output[0].w, temp[0].___w, temp[1].___w;
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: TEX temp[0], input[2].xy__, 2D[0];
>  1: TEX temp[2].yw, input[2].xy__, 2D[1];
>  2: MOV temp[2].x, temp[2].w___;
>  3: MAD temp[2].xy, temp[2].xy__, const[7].xx__, none.-1-1__;
>  4: MOV temp[2].z, none.__0_;
>  5: DP3 temp[3].x, temp[2].xyz_, temp[2].xyz_;
>  6: ADD temp[3].x, none.1___, -temp[3].x___;
>  7: RSQ temp[3].x, |temp[3].x___|;
>  8: RCP temp[2].z, temp[3].x___;
>  9: DP3 temp[4].x, input[3].xyz_, temp[2].xyz_;
> 10: DP3 temp[4].y, input[4].xyz_, temp[2].xyz_;
> 11: DP3 temp[4].z, input[5].xyz_, temp[2].xyz_;
> 12: TEX temp[5].xyz, temp[4].xyz_, CUBE[2];
> 13: MUL temp[5].xyz, temp[5].xyz_, const[3].xyz_;
> 14: MUL temp[5].xyz, temp[5].xyz_, const[7].xxx_;
> 15: MUL temp[6].xyz, temp[0].xyz_, const[4].xyz_;
> 16: MUL temp[6].xyz, temp[5].xyz_, temp[6].xyz_;
> 17: DP3 temp[3].x, input[1].xyz_, temp[4].xyz_;
> 18: ADD temp[3].x, temp[3].x___, temp[3].x___;
> 19: MAD temp[7].xyz, temp[4].xyz_, temp[3].xxx_, -input[1].xyz_;
> 20: TEX temp[8].xyz, temp[7].xyz_, CUBE[3];
> 21: TEX temp[9].xyz, input[2].xy__, 2D[4];
> 22: ADD temp[1].xyz, temp[9].xyz_, temp[9].xyz_;
> 23: MAD temp[6].xyz, temp[1].xyz_, temp[8].xyz_, temp[6].xyz_;
> 24: MUL temp[3].xyz, temp[6].xyz_, input[0].xyz_;
> 25: DP3 temp[1], input[1].xyz_, temp[4].xyz_;
> 26: ADD temp[1], none.1111, -temp[1];
> 27: MUL temp[1], temp[1], temp[1];
> 28: MUL temp[1], temp[1], temp[1];
> 29: MUL temp[10], temp[1], const[5];
> 30: MUL temp[11].xyz, temp[10].xyz_, temp[5].xyz_;
> 31: MAD temp[11].xyz, temp[10].xyz_, const[6].www_, temp[11].xyz_;
> 32: ADD output[0].xyz, temp[3].xyz_, temp[11].xyz_;
> 33: ADD temp[1].w, none.___1, -temp[10].___w;
> 34: MUL output[0].w, temp[0].___w, temp[1].___w;
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: TEX temp[0], input[2].xy__, 2D[0];
>  1: TEX temp[2].yw, input[2].xy__, 2D[1];
>  2: MOV temp[2].x, temp[2].w___;
>  3: MAD temp[2].xy, temp[2].xy__, const[7].xx__, none.-1-1__;
>  4: MOV temp[2].z, none.__0_;
>  5: DP3 temp[3].x, temp[2].xyz_, temp[2].xyz_;
>  6: ADD temp[3].x, none.1___, -temp[3].x___;
>  7: RSQ temp[3].x, |temp[3].x___|;
>  8: RCP temp[2].z, temp[3].x___;
>  9: DP3 temp[4].x, input[3].xyz_, temp[2].xyz_;
> 10: DP3 temp[4].y, input[4].xyz_, temp[2].xyz_;
> 11: DP3 temp[4].z, input[5].xyz_, temp[2].xyz_;
> 12: TEX temp[5].xyz, temp[4].xyz_, CUBE[2];
> 13: MUL temp[5].xyz, temp[5].xyz_, const[3].xyz_;
> 14: MUL temp[5].xyz, temp[5].xyz_, const[7].xxx_;
> 15: MUL temp[6].xyz, temp[0].xyz_, const[4].xyz_;
> 16: MUL temp[6].xyz, temp[5].xyz_, temp[6].xyz_;
> 17: DP3 temp[3].x, input[1].xyz_, temp[4].xyz_;
> 18: ADD temp[3].x, temp[3].x___, temp[3].x___;
> 19: MAD temp[7].xyz, temp[4].xyz_, temp[3].xxx_, -input[1].xyz_;
> 20: TEX temp[8].xyz, temp[7].xyz_, CUBE[3];
> 21: TEX temp[9].xyz, input[2].xy__, 2D[4];
> 22: ADD temp[1].xyz, temp[9].xyz_, temp[9].xyz_;
> 23: MAD temp[6].xyz, temp[1].xyz_, temp[8].xyz_, temp[6].xyz_;
> 24: MUL temp[3].xyz, temp[6].xyz_, input[0].xyz_;
> 25: DP3 temp[1], input[1].xyz_, temp[4].xyz_;
> 26: ADD temp[1], none.1111, -temp[1];
> 27: MUL temp[1], temp[1], temp[1];
> 28: MUL temp[1], temp[1], temp[1];
> 29: MUL temp[10], temp[1], const[5];
> 30: MUL temp[11].xyz, temp[10].xyz_, temp[5].xyz_;
> 31: MAD temp[11].xyz, temp[10].xyz_, const[6].www_, temp[11].xyz_;
> 32: ADD output[0].xyz, temp[3].xyz_, temp[11].xyz_;
> 33: ADD temp[1].w, none.___1, -temp[10].___w;
> 34: MUL output[0].w, temp[0].___w, temp[1].___w;
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: TEX temp[0], input[2].xy__, 2D[0];
>  1: TEX temp[2].yw, input[2].xy__, 2D[1];
>  2: src0.w = temp[2]
>     MAD temp[2].x, src0.w__, src0.111, src0.000
>  3: src0.xyz = temp[2], src1.xyz = const[7]
>     MAD temp[2].xy, src0.xy_, src1.xx_, -src0.11_
>  4: 
>     MAD temp[2].z, src0.__0, src0.111, src0.000
>  5: src0.xyz = temp[2]
>     DP3 temp[3].x, src0.xyz, src0.xyz
>  6: src0.xyz = temp[3]
>     MAD temp[3].x, src0.1__, src0.111, -src0.x__
>  7: src0.xyz = temp[3]
>     REPL_ALPHA temp[3].x
>     RSQ, |src0.x|
>  8: src0.xyz = temp[3]
>     REPL_ALPHA temp[2].z
>     RCP, src0.x
>  9: src0.xyz = input[3], src1.xyz = temp[2]
>     DP3 temp[4].x, src0.xyz, src1.xyz
> 10: src0.xyz = input[4], src1.xyz = temp[2]
>     DP3 temp[4].y, src0.xyz, src1.xyz
> 11: src0.xyz = input[5], src1.xyz = temp[2]
>     DP3 temp[4].z, src0.xyz, src1.xyz
> 12: TEX temp[5].xyz, temp[4].xyz_, CUBE[2];
> 13: src0.xyz = temp[5], src1.xyz = const[3]
>     MAD temp[5].xyz, src0.xyz, src1.xyz, src0.000
> 14: src0.xyz = temp[5], src1.xyz = const[7]
>     MAD temp[5].xyz, src0.xyz, src1.xxx, src0.000
> 15: src0.xyz = temp[0], src1.xyz = const[4]
>     MAD temp[6].xyz, src0.xyz, src1.xyz, src0.000
> 16: src0.xyz = temp[5], src1.xyz = temp[6]
>     MAD temp[6].xyz, src0.xyz, src1.xyz, src0.000
> 17: src0.xyz = input[1], src1.xyz = temp[4]
>     DP3 temp[3].x, src0.xyz, src1.xyz
> 18: src0.xyz = temp[3]
>     MAD temp[3].x, src0.x__, src0.111, src0.x__
> 19: src0.xyz = temp[4], src1.xyz = temp[3], src2.xyz = input[1]
>     MAD temp[7].xyz, src0.xyz, src1.xxx, -src2.xyz
> 20: TEX temp[8].xyz, temp[7].xyz_, CUBE[3];
> 21: TEX temp[9].xyz, input[2].xy__, 2D[4];
> 22: src0.xyz = temp[9]
>     MAD temp[1].xyz, src0.xyz, src0.111, src0.xyz
> 23: src0.xyz = temp[1], src1.xyz = temp[8], src2.xyz = temp[6]
>     MAD temp[6].xyz, src0.xyz, src1.xyz, src2.xyz
> 24: src0.xyz = temp[6], src1.xyz = input[0]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src0.000
> 25: src0.xyz = input[1], src1.xyz = temp[4]
>     DP3 temp[1].xyz, src0.xyz, src1.xyz
>     DP3 temp[1].w, src0._, src0._
> 26: src0.xyz = temp[1], src0.w = temp[1]
>     MAD temp[1].xyz, src0.111, src0.111, -src0.xyz
>     MAD temp[1].w, src0.1, src0.1, -src0.w
> 27: src0.xyz = temp[1], src0.w = temp[1]
>     MAD temp[1].xyz, src0.xyz, src0.xyz, src0.000
>     MAD temp[1].w, src0.w, src0.w, src0.0
> 28: src0.xyz = temp[1], src0.w = temp[1]
>     MAD temp[1].xyz, src0.xyz, src0.xyz, src0.000
>     MAD temp[1].w, src0.w, src0.w, src0.0
> 29: src0.xyz = temp[1], src0.w = temp[1], src1.xyz = const[5], src1.w = const[5]
>     MAD temp[10].xyz, src0.xyz, src1.xyz, src0.000
>     MAD temp[10].w, src0.w, src1.w, src0.0
> 30: src0.xyz = temp[10], src1.xyz = temp[5]
>     MAD temp[11].xyz, src0.xyz, src1.xyz, src0.000
> 31: src0.xyz = temp[10], src0.w = const[6], src1.xyz = temp[11]
>     MAD temp[11].xyz, src0.xyz, src0.www, src1.xyz
> 32: src0.xyz = temp[3], src1.xyz = temp[11]
>     MAD color[0].xyz, src0.xyz, src0.111, src1.xyz
> 33: src0.w = temp[10]
>     MAD temp[1].w, src0.1, src0.1, -src0.w
> 34: src0.w = temp[0], src1.w = temp[1]
>     MAD color[0].w, src0.w, src1.w, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[9].xyz, input[2].xy__, 2D[4];
>  2: TEX temp[2].yw, input[2].xy__, 2D[1];
>  3: TEX temp[0], input[2].xy__, 2D[0];
>  4: src0.xyz = temp[0], src1.xyz = const[4]
>     MAD temp[6].xyz, src0.xyz, src1.xyz, src0.000
>  5: src0.w = temp[2]
>     MAD temp[2].x, src0.w__, src0.111, src0.000
>  6: src0.xyz = temp[2], src1.xyz = const[7]
>     MAD temp[2].xy, src0.xy_, src1.xx_, -src0.11_
>  7: src0.xyz = temp[9]
>     MAD temp[1].xyz, src0.xyz, src0.111, src0.xyz
>  8: 
>     MAD temp[2].z, src0.__0, src0.111, src0.000
>  9: src0.xyz = temp[2]
>     DP3 temp[3].x, src0.xyz, src0.xyz
> 10: src0.xyz = temp[3]
>     MAD temp[3].x, src0.1__, src0.111, -src0.x__
> 11: src0.xyz = temp[3]
>     REPL_ALPHA temp[3].x
>     RSQ, |src0.x|
> 12: src0.xyz = temp[3]
>     REPL_ALPHA temp[2].z
>     RCP, src0.x
> 13: src0.xyz = input[3], src1.xyz = temp[2]
>     DP3 temp[4].x, src0.xyz, src1.xyz
> 14: src0.xyz = input[4], src1.xyz = temp[2]
>     DP3 temp[4].y, src0.xyz, src1.xyz
> 15: src0.xyz = input[5], src1.xyz = temp[2]
>     DP3 temp[4].z, src0.xyz, src1.xyz
> 16: src0.xyz = input[1], src1.xyz = temp[4]
>     DP3 temp[3].x, src0.xyz, src1.xyz
> 17: src0.xyz = temp[3]
>     MAD temp[3].x, src0.x__, src0.111, src0.x__
> 18: src0.xyz = temp[4], src1.xyz = temp[3], src2.xyz = input[1]
>     MAD temp[7].xyz, src0.xyz, src1.xxx, -src2.xyz
> 19: BEGIN_TEX;
> 20: TEX temp[8].xyz, temp[7].xyz_, CUBE[3];
> 21: TEX temp[5].xyz, temp[4].xyz_, CUBE[2];
> 22: src0.xyz = temp[5], src1.xyz = const[3]
>     MAD temp[5].xyz, src0.xyz, src1.xyz, src0.000
> 23: src0.xyz = temp[5], src1.xyz = const[7]
>     MAD temp[5].xyz, src0.xyz, src1.xxx, src0.000
> 24: src0.xyz = temp[5], src1.xyz = temp[6]
>     MAD temp[6].xyz, src0.xyz, src1.xyz, src0.000
> 25: src0.xyz = temp[1], src1.xyz = temp[8], src2.xyz = temp[6]
>     MAD temp[6].xyz, src0.xyz, src1.xyz, src2.xyz
> 26: src0.xyz = input[1], src1.xyz = temp[4]
>     DP3 temp[1].xyz, src0.xyz, src1.xyz
>     DP3 temp[1].w, src0._, src0._
> 27: src0.xyz = temp[1], src0.w = temp[1]
>     MAD temp[1].xyz, src0.111, src0.111, -src0.xyz
>     MAD temp[1].w, src0.1, src0.1, -src0.w
> 28: src0.xyz = temp[1], src0.w = temp[1]
>     MAD temp[1].xyz, src0.xyz, src0.xyz, src0.000
>     MAD temp[1].w, src0.w, src0.w, src0.0
> 29: src0.xyz = temp[1], src0.w = temp[1]
>     MAD temp[1].xyz, src0.xyz, src0.xyz, src0.000
>     MAD temp[1].w, src0.w, src0.w, src0.0
> 30: src0.xyz = temp[1], src0.w = temp[1], src1.xyz = const[5], src1.w = const[5]
>     MAD temp[10].xyz, src0.xyz, src1.xyz, src0.000
>     MAD temp[10].w, src0.w, src1.w, src0.0
> 31: src0.xyz = temp[10], src0.w = temp[10], src1.xyz = temp[5]
>     MAD temp[11].xyz, src0.xyz, src1.xyz, src0.000
>     MAD temp[1].w, src0.1, src0.1, -src0.w
> 32: src0.xyz = temp[10], src0.w = const[6], src1.xyz = temp[11], src1.w = temp[0], src2.w = temp[1]
>     MAD temp[11].xyz, src0.xyz, src0.www, src1.xyz
>     MAD color[0].w, src1.w, src2.w, src0.0
> 33: src0.xyz = temp[6], src1.xyz = input[0]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src0.000
> 34: src0.xyz = temp[3], src1.xyz = temp[11]
>     MAD color[0].xyz, src0.xyz, src0.111, src1.xyz
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[6].xyz, temp[2].xy__, 2D[4];
>  2: TEX temp[7].yw, temp[2].xy__, 2D[1];
>  3: TEX temp[2], temp[2].xy__, 2D[0];
>  4: src0.xyz = temp[2], src1.xyz = const[4]
>     MAD temp[9].xyz, src0.xyz, src1.xyz, src0.000
>  5: src0.w = temp[7]
>     MAD temp[7].x, src0.w__, src0.111, src0.000
>  6: src0.xyz = temp[7], src1.xyz = const[7]
>     MAD temp[7].xy, src0.xy_, src1.xx_, -src0.11_
>  7: src0.xyz = temp[6]
>     MAD temp[6].xyz, src0.xyz, src0.111, src0.xyz
>  8: 
>     MAD temp[7].z, src0.__0, src0.111, src0.000
>  9: src0.xyz = temp[7]
>     DP3 temp[8].x, src0.xyz, src0.xyz
> 10: src0.xyz = temp[8]
>     MAD temp[8].x, src0.1__, src0.111, -src0.x__
> 11: src0.xyz = temp[8]
>     REPL_ALPHA temp[8].x
>     RSQ, |src0.x|
> 12: src0.xyz = temp[8]
>     REPL_ALPHA temp[7].z
>     RCP, src0.x
> 13: src0.xyz = temp[3], src1.xyz = temp[7]
>     DP3 temp[3].x, src0.xyz, src1.xyz
> 14: src0.xyz = temp[4], src1.xyz = temp[7]
>     DP3 temp[3].y, src0.xyz, src1.xyz
> 15: src0.xyz = temp[5], src1.xyz = temp[7]
>     DP3 temp[3].z, src0.xyz, src1.xyz
> 16: src0.xyz = temp[1], src1.xyz = temp[3]
>     DP3 temp[8].x, src0.xyz, src1.xyz
> 17: src0.xyz = temp[8]
>     MAD temp[8].x, src0.x__, src0.111, src0.x__
> 18: src0.xyz = temp[3], src1.xyz = temp[8], src2.xyz = temp[1]
>     MAD temp[4].xyz, src0.xyz, src1.xxx, -src2.xyz
> 19: BEGIN_TEX;
> 20: TEX temp[5].xyz, temp[4].xyz_, CUBE[3];
> 21: TEX temp[4].xyz, temp[3].xyz_, CUBE[2];
> 22: src0.xyz = temp[4], src1.xyz = const[3]
>     MAD temp[4].xyz, src0.xyz, src1.xyz, src0.000
> 23: src0.xyz = temp[4], src1.xyz = const[7]
>     MAD temp[4].xyz, src0.xyz, src1.xxx, src0.000
> 24: src0.xyz = temp[4], src1.xyz = temp[9]
>     MAD temp[9].xyz, src0.xyz, src1.xyz, src0.000
> 25: src0.xyz = temp[6], src1.xyz = temp[5], src2.xyz = temp[9]
>     MAD temp[9].xyz, src0.xyz, src1.xyz, src2.xyz
> 26: src0.xyz = temp[1], src1.xyz = temp[3]
>     DP3 temp[6].xyz, src0.xyz, src1.xyz
>     DP3 temp[6].w, src0._, src0._
> 27: src0.xyz = temp[6], src0.w = temp[6]
>     MAD temp[6].xyz, src0.111, src0.111, -src0.xyz
>     MAD temp[6].w, src0.1, src0.1, -src0.w
> 28: src0.xyz = temp[6], src0.w = temp[6]
>     MAD temp[6].xyz, src0.xyz, src0.xyz, src0.000
>     MAD temp[6].w, src0.w, src0.w, src0.0
> 29: src0.xyz = temp[6], src0.w = temp[6]
>     MAD temp[6].xyz, src0.xyz, src0.xyz, src0.000
>     MAD temp[6].w, src0.w, src0.w, src0.0
> 30: src0.xyz = temp[6], src0.w = temp[6], src1.xyz = const[5], src1.w = const[5]
>     MAD temp[1].xyz, src0.xyz, src1.xyz, src0.000
>     MAD temp[1].w, src0.w, src1.w, src0.0
> 31: src0.xyz = temp[1], src0.w = temp[1], src1.xyz = temp[4]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src0.000
>     MAD temp[6].w, src0.1, src0.1, -src0.w
> 32: src0.xyz = temp[1], src0.w = const[6], src1.xyz = temp[3], src1.w = temp[2], src2.w = temp[6]
>     MAD temp[3].xyz, src0.xyz, src0.www, src1.xyz
>     MAD color[0].w, src1.w, src2.w, src0.0
> 33: src0.xyz = temp[9], src1.xyz = temp[0]
>     MAD temp[8].xyz, src0.xyz, src1.xyz, src0.000
> 34: src0.xyz = temp[8], src1.xyz = temp[3]
>     MAD color[0].xyz, src0.xyz, src0.111, src1.xyz
>pc=7*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 14, tex_end: 2  (code_addr: 00040380)
>  TEX:
>    TEX t6, t2, texture[4]   (0000a182)
>    TEX t7, t2, texture[1]   (000089c2)
>    TEX t2, t2, texture[0]   (00008082)
>  0: xyz:  t2  c4  t0 -> t9.xyz               (03a40902)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz   c4.xyz      0.0    op: 00050200
>       w:     t2.x     t2.x     t2.x    op: 00000000
>  1: xyz:  t0  t0  t0 -> t7.x                 (009c0000)
>       w:  t7  t0  t0 ->                      (00000007)
>     xyz:   t7.www      1.0      0.0    op: 00050a8c
>       w:     t0.x     t0.x     t0.x    op: 00000000
>  2: xyz:  t7  c7  t0 -> t7.xy                (019c09c7)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t7.xyz   c7.xxx     -1.0    op: 000d4280
>       w:     t7.x     t7.x     t7.x    op: 00000000
>  3: xyz:  t6  t0  t0 -> t6.xyz               (03980006)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t6.xyz      1.0   t6.xyz    op: 00000a80
>       w:     t6.x     t6.x     t6.x    op: 00000000
>  4: xyz:  t0  t0  t0 -> t7.z                 (021c0000)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:      0.0      1.0      0.0    op: 00050a94
>       w:     t0.x     t0.x     t0.x    op: 00000000
>  5: xyz:  t7  t0  t0 -> t8.x                 (00a00007)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t7.xyz   t7.xyz   t7.xxx    op: 00804000
>       w:     t7.x     t7.x     t7.x    op: 00000000
>  6: xyz:  t8  t0  t0 -> t8.x                 (00a00008)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:      1.0      1.0  -t8.xyz    op: 00080a95
>       w:     t8.x     t8.x     t8.x    op: 00000000
>  7: xyz:  t8  t0  t0 -> t8.x                 (00a00008)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t8.xxx   t8.xxx   t8.xxx    op: 05004081
>       w:   |t8.x|     t8.x     t8.x    op: 05800040
>  8: xyz:  t8  t0  t0 -> t7.z                 (021c0008)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t8.xxx   t8.xxx   t8.xxx    op: 05004081
>       w:     t8.x     t8.x     t8.x    op: 05000000
>  9: xyz:  t3  t7  t0 -> t3.x                 (008c01c3)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t3.xyz   t7.xyz   t3.xxx    op: 00804200
>       w:     t3.x     t3.x     t3.x    op: 00000000
> 10: xyz:  t4  t7  t0 -> t3.y                 (010c01c4)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t4.xyz   t7.xyz   t4.xxx    op: 00804200
>       w:     t4.x     t4.x     t4.x    op: 00000000
> 11: xyz:  t5  t7  t0 -> t3.z                 (020c01c5)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t5.xyz   t7.xyz   t5.xxx    op: 00804200
>       w:     t5.x     t5.x     t5.x    op: 00000000
> 12: xyz:  t1  t3  t0 -> t8.x                 (00a000c1)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   t3.xyz   t1.xxx    op: 00804200
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 13: xyz:  t8  t0  t0 -> t8.x                 (00a00008)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t8.xyz      1.0   t8.xyz    op: 00000a80
>       w:     t8.x     t8.x     t8.x    op: 00000000
> 14: xyz:  t3  t8  t1 -> t4.xyz               (03901203)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t3.xyz   t8.xxx  -t1.xyz    op: 000a0280
>       w:     t3.x     t3.x     t3.x    op: 00000000
>NODE 1: alu_offset: 15, tex_offset: 3, alu_end: 12, tex_end: 1  (code_addr: 0042330f)
>  TEX:
>    TEX t5, t4, texture[3]   (00009944)
>    TEX t4, t3, texture[2]   (00009103)
> 15: xyz:  t4  c3  t0 -> t4.xyz               (039008c4)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t4.xyz   c3.xyz      0.0    op: 00050200
>       w:     t4.x     t4.x     t4.x    op: 00000000
> 16: xyz:  t4  c7  t0 -> t4.xyz               (039009c4)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t4.xyz   c7.xxx      0.0    op: 00050280
>       w:     t4.x     t4.x     t4.x    op: 00000000
> 17: xyz:  t4  t9  t0 -> t9.xyz               (03a40244)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t4.xyz   t9.xyz      0.0    op: 00050200
>       w:     t4.x     t4.x     t4.x    op: 00000000
> 18: xyz:  t6  t5  t9 -> t9.xyz               (03a49146)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t6.xyz   t5.xyz   t9.xyz    op: 00020200
>       w:     t6.x     t6.x     t6.x    op: 00000000
> 19: xyz:  t1  t3  t0 -> t6.xyz               (039800c1)
>       w:  t0  t0  t0 -> t6.w                 (00980000)
>     xyz:   t1.xyz   t3.xyz   t1.xxx    op: 00804200
>       w:      1.0      1.0     t1.x    op: 00800891
> 20: xyz:  t6  t0  t0 -> t6.xyz               (03980006)
>       w:  t6  t0  t0 -> t6.w                 (00980006)
>     xyz:      1.0      1.0  -t6.xyz    op: 00080a95
>       w:      1.0      1.0    -t6.w    op: 000a4891
> 21: xyz:  t6  t0  t0 -> t6.xyz               (03980006)
>       w:  t6  t0  t0 -> t6.w                 (00980006)
>     xyz:   t6.xyz   t6.xyz      0.0    op: 00050000
>       w:     t6.w     t6.w      0.0    op: 00040489
> 22: xyz:  t6  t0  t0 -> t6.xyz               (03980006)
>       w:  t6  t0  t0 -> t6.w                 (00980006)
>     xyz:   t6.xyz   t6.xyz      0.0    op: 00050000
>       w:     t6.w     t6.w      0.0    op: 00040489
> 23: xyz:  t6  c5  t0 -> t1.xyz               (03840946)
>       w:  t6  c5  t0 -> t1.w                 (00840946)
>     xyz:   t6.xyz   c5.xyz      0.0    op: 00050200
>       w:     t6.w     c5.w      0.0    op: 00040509
> 24: xyz:  t1  t4  t0 -> t3.xyz               (038c0101)
>       w:  t1  t0  t0 -> t6.w                 (00980001)
>     xyz:   t1.xyz   t4.xyz      0.0    op: 00050200
>       w:      1.0      1.0    -t1.w    op: 000a4891
> 25: xyz:  t1  t3  t0 -> t3.xyz               (038c00c1)
>       w:  c6  t2  t6 -> o0.w                 (010060a6)
>     xyz:   t1.xyz   c6.www   t3.xyz    op: 00010600
>       w:     t2.w     t6.w      0.0    op: 0004058a
> 26: xyz:  t9  t0  t0 -> t8.xyz               (03a00009)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t9.xyz   t0.xyz      0.0    op: 00050200
>       w:     t9.x     t9.x     t9.x    op: 00000000
> 27: xyz:  t8  t3  t0 -> o0.xyz               (1c0000c8)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t8.xyz      1.0   t3.xyz    op: 00010a80
>       w:     t8.x     t8.x     t8.x    op: 00000000
>r300: Initial fragment program
>FRAG
>DCL IN[0], GENERIC[1], PERSPECTIVE
>DCL IN[1], GENERIC[2], PERSPECTIVE
>DCL IN[2], GENERIC[5], PERSPECTIVE
>DCL IN[3], GENERIC[6], PERSPECTIVE
>DCL IN[4], GENERIC[7], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL CONST[3]
>DCL TEMP[0..3]
>IMM FLT32 {    2.0000,    -1.0000,     0.0000,     1.0000}
>IMM FLT32 {    0.0000,     1.0000,     5.0000,     0.0000}
>  0: TEX TEMP[0], IN[0], SAMP[0], 2D
>  1: MOV TEMP[0].x, TEMP[0].wwww
>  2: MAD TEMP[0].xy, TEMP[0], IMM[0].xxxx, IMM[0].yyyy
>  3: MOV TEMP[0].z, IMM[1].xxxx
>  4: DP3 TEMP[1].x, TEMP[0], TEMP[0]
>  5: ADD TEMP[1].x, IMM[1].yyyy, -TEMP[1].xxxx
>  6: RSQ TEMP[1].x, TEMP[1].xxxx
>  7: RCP TEMP[0].z, TEMP[1].xxxx
>  8: DP3 TEMP[2].x, IN[2], TEMP[0]
>  9: DP3 TEMP[2].y, IN[3], TEMP[0]
> 10: DP3 TEMP[2].z, IN[4], TEMP[0]
> 11: DP3 TEMP[3], IN[1], TEMP[2]
> 12: MUL TEMP[3], TEMP[3], TEMP[3]
> 13: MUL_SAT TEMP[3], TEMP[3], IMM[1].zzzz
> 14: MUL TEMP[3], TEMP[3], CONST[3]
> 15: MOV OUT[0].xyz, TEMP[3]
> 16: MOV OUT[0].w, IMM[1].yyyy
> 17: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: TEX temp[0], input[0], 2D[0];
>  1: MOV temp[0].x, temp[0].wwww;
>  2: MAD temp[0].xy, temp[0], const[4].xxxx, const[4].yyyy;
>  3: MOV temp[0].z, const[5].xxxx;
>  4: DP3 temp[1].x, temp[0], temp[0];
>  5: ADD temp[1].x, const[5].yyyy, -temp[1].xxxx;
>  6: RSQ temp[1].x, temp[1].xxxx;
>  7: RCP temp[0].z, temp[1].xxxx;
>  8: DP3 temp[2].x, input[2], temp[0];
>  9: DP3 temp[2].y, input[3], temp[0];
> 10: DP3 temp[2].z, input[4], temp[0];
> 11: DP3 temp[3], input[1], temp[2];
> 12: MUL temp[3], temp[3], temp[3];
> 13: MUL_SAT temp[3], temp[3], const[5].zzzz;
> 14: MUL temp[3], temp[3], const[3];
> 15: MOV output[0].xyz, temp[3];
> 16: MOV output[0].w, const[5].yyyy;
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: TEX temp[0], input[0], 2D[0];
>  1: MOV temp[0].x, temp[0].wwww;
>  2: MAD temp[0].xy, temp[0], const[4].xxxx, const[4].yyyy;
>  3: MOV temp[0].z, const[5].xxxx;
>  4: DP3 temp[1].x, temp[0], temp[0];
>  5: ADD temp[1].x, const[5].yyyy, -temp[1].xxxx;
>  6: RSQ temp[1].x, temp[1].xxxx;
>  7: RCP temp[0].z, temp[1].xxxx;
>  8: DP3 temp[2].x, input[2], temp[0];
>  9: DP3 temp[2].y, input[3], temp[0];
> 10: DP3 temp[2].z, input[4], temp[0];
> 11: DP3 temp[3], input[1], temp[2];
> 12: MUL temp[3], temp[3], temp[3];
> 13: MUL_SAT temp[3], temp[3], const[5].zzzz;
> 14: MUL temp[3], temp[3], const[3];
> 15: MOV output[0].xyz, temp[3];
> 16: MOV output[0].w, const[5].yyyy;
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: TEX temp[0], input[0], 2D[0];
>  1: MOV temp[0].x, temp[0].wwww;
>  2: MAD temp[0].xy, temp[0], const[4].xxxx, const[4].yyyy;
>  3: MOV temp[0].z, const[5].xxxx;
>  4: DP3 temp[1].x, temp[0], temp[0];
>  5: ADD temp[1].x, const[5].yyyy, -temp[1].xxxx;
>  6: RSQ temp[1].x, temp[1].xxxx;
>  7: RCP temp[0].z, temp[1].xxxx;
>  8: DP3 temp[2].x, input[2], temp[0];
>  9: DP3 temp[2].y, input[3], temp[0];
> 10: DP3 temp[2].z, input[4], temp[0];
> 11: DP3 temp[3], input[1], temp[2];
> 12: MUL temp[3], temp[3], temp[3];
> 13: MUL_SAT temp[3], temp[3], const[5].zzzz;
> 14: MUL temp[3], temp[3], const[3];
> 15: MOV output[0].xyz, temp[3];
> 16: MOV output[0].w, const[5].yyyy;
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: TEX temp[0], input[0], 2D[0];
>  1: MOV temp[0].x, temp[0].wwww;
>  2: MAD temp[0].xy, temp[0], const[4].xxxx, const[4].yyyy;
>  3: MOV temp[0].z, const[5].xxxx;
>  4: DP3 temp[1].x, temp[0], temp[0];
>  5: ADD temp[1].x, const[5].yyyy, -temp[1].xxxx;
>  6: RSQ temp[1].x, |temp[1].xxxx|;
>  7: RCP temp[0].z, temp[1].xxxx;
>  8: DP3 temp[2].x, input[2], temp[0];
>  9: DP3 temp[2].y, input[3], temp[0];
> 10: DP3 temp[2].z, input[4], temp[0];
> 11: DP3 temp[3], input[1], temp[2];
> 12: MUL temp[3], temp[3], temp[3];
> 13: MUL_SAT temp[3], temp[3], const[5].zzzz;
> 14: MUL temp[3], temp[3], const[3];
> 15: MOV output[0].xyz, temp[3];
> 16: MOV output[0].w, const[5].yyyy;
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: TEX temp[0], input[0], 2D[0];
>  1: MOV temp[0].x, temp[0].wwww;
>  2: MAD temp[0].xy, temp[0], const[4].xxxx, const[4].yyyy;
>  3: MOV temp[0].z, const[5].xxxx;
>  4: DP3 temp[1].x, temp[0], temp[0];
>  5: ADD temp[1].x, const[5].yyyy, -temp[1].xxxx;
>  6: RSQ temp[1].x, |temp[1].xxxx|;
>  7: RCP temp[0].z, temp[1].xxxx;
>  8: DP3 temp[2].x, input[2], temp[0];
>  9: DP3 temp[2].y, input[3], temp[0];
> 10: DP3 temp[2].z, input[4], temp[0];
> 11: DP3 temp[3], input[1], temp[2];
> 12: MUL temp[3], temp[3], temp[3];
> 13: MUL_SAT temp[3], temp[3], const[5].zzzz;
> 14: MUL temp[3], temp[3], const[3];
> 15: MOV output[0].xyz, temp[3];
> 16: MOV output[0].w, const[5].yyyy;
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: TEX temp[0].yw, input[0].xy__, 2D[0];
>  1: MOV temp[0].x, temp[0].w___;
>  2: MAD temp[0].xy, temp[0].xy__, const[4].xx__, const[4].yy__;
>  3: MOV temp[0].z, const[5].__x_;
>  4: DP3 temp[1].x, temp[0].xyz_, temp[0].xyz_;
>  5: ADD temp[1].x, const[5].y___, -temp[1].x___;
>  6: RSQ temp[1].x, |temp[1].x___|;
>  7: RCP temp[0].z, temp[1].x___;
>  8: DP3 temp[2].x, input[2].xyz_, temp[0].xyz_;
>  9: DP3 temp[2].y, input[3].xyz_, temp[0].xyz_;
> 10: DP3 temp[2].z, input[4].xyz_, temp[0].xyz_;
> 11: DP3 temp[3].xyz, input[1].xyz_, temp[2].xyz_;
> 12: MUL temp[3].xyz, temp[3].xyz_, temp[3].xyz_;
> 13: MUL_SAT temp[3].xyz, temp[3].xyz_, const[5].zzz_;
> 14: MUL temp[3].xyz, temp[3].xyz_, const[3].xyz_;
> 15: MOV output[0].xyz, temp[3].xyz_;
> 16: MOV output[0].w, const[5].___y;
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: TEX temp[0].yw, input[0].xy__, 2D[0];
>  1: MOV temp[0].x, temp[0].w___;
>  2: MAD temp[0].xy, temp[0].xy__, const[4].xx__, none.-1-1__;
>  3: MOV temp[0].z, none.__0_;
>  4: DP3 temp[1].x, temp[0].xyz_, temp[0].xyz_;
>  5: ADD temp[1].x, none.1___, -temp[1].x___;
>  6: RSQ temp[1].x, |temp[1].x___|;
>  7: RCP temp[0].z, temp[1].x___;
>  8: DP3 temp[2].x, input[2].xyz_, temp[0].xyz_;
>  9: DP3 temp[2].y, input[3].xyz_, temp[0].xyz_;
> 10: DP3 temp[2].z, input[4].xyz_, temp[0].xyz_;
> 11: DP3 temp[3].xyz, input[1].xyz_, temp[2].xyz_;
> 12: MUL temp[3].xyz, temp[3].xyz_, temp[3].xyz_;
> 13: MUL_SAT temp[3].xyz, temp[3].xyz_, const[5].zzz_;
> 14: MUL temp[3].xyz, temp[3].xyz_, const[3].xyz_;
> 15: MOV output[0].xyz, temp[3].xyz_;
> 16: MOV output[0].w, none.___1;
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: TEX temp[0].yw, input[0].xy__, 2D[0];
>  1: MOV temp[0].x, temp[0].w___;
>  2: MAD temp[0].xy, temp[0].xy__, const[4].xx__, none.-1-1__;
>  3: MOV temp[0].z, none.__0_;
>  4: DP3 temp[1].x, temp[0].xyz_, temp[0].xyz_;
>  5: ADD temp[1].x, none.1___, -temp[1].x___;
>  6: RSQ temp[1].x, |temp[1].x___|;
>  7: RCP temp[0].z, temp[1].x___;
>  8: DP3 temp[2].x, input[2].xyz_, temp[0].xyz_;
>  9: DP3 temp[2].y, input[3].xyz_, temp[0].xyz_;
> 10: DP3 temp[2].z, input[4].xyz_, temp[0].xyz_;
> 11: DP3 temp[3].xyz, input[1].xyz_, temp[2].xyz_;
> 12: MUL temp[3].xyz, temp[3].xyz_, temp[3].xyz_;
> 13: MUL_SAT temp[3].xyz, temp[3].xyz_, const[5].zzz_;
> 14: MUL temp[3].xyz, temp[3].xyz_, const[3].xyz_;
> 15: MOV output[0].xyz, temp[3].xyz_;
> 16: MOV output[0].w, none.___1;
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: TEX temp[0].yw, input[0].xy__, 2D[0];
>  1: src0.w = temp[0]
>     MAD temp[0].x, src0.w__, src0.111, src0.000
>  2: src0.xyz = temp[0], src1.xyz = const[4]
>     MAD temp[0].xy, src0.xy_, src1.xx_, -src0.11_
>  3: 
>     MAD temp[0].z, src0.__0, src0.111, src0.000
>  4: src0.xyz = temp[0]
>     DP3 temp[1].x, src0.xyz, src0.xyz
>  5: src0.xyz = temp[1]
>     MAD temp[1].x, src0.1__, src0.111, -src0.x__
>  6: src0.xyz = temp[1]
>     REPL_ALPHA temp[1].x
>     RSQ, |src0.x|
>  7: src0.xyz = temp[1]
>     REPL_ALPHA temp[0].z
>     RCP, src0.x
>  8: src0.xyz = input[2], src1.xyz = temp[0]
>     DP3 temp[2].x, src0.xyz, src1.xyz
>  9: src0.xyz = input[3], src1.xyz = temp[0]
>     DP3 temp[2].y, src0.xyz, src1.xyz
> 10: src0.xyz = input[4], src1.xyz = temp[0]
>     DP3 temp[2].z, src0.xyz, src1.xyz
> 11: src0.xyz = input[1], src1.xyz = temp[2]
>     DP3 temp[3].xyz, src0.xyz, src1.xyz
> 12: src0.xyz = temp[3]
>     MAD temp[3].xyz, src0.xyz, src0.xyz, src0.000
> 13: src0.xyz = temp[3], src1.xyz = const[5]
>     MAD_SAT temp[3].xyz, src0.xyz, src1.zzz, src0.000
> 14: src0.xyz = temp[3], src1.xyz = const[3]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src0.000
> 15: src0.xyz = temp[3]
>     MAD color[0].xyz, src0.xyz, src0.111, src0.000
> 16: 
>     MAD color[0].w, src0.1, src0.1, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[0].yw, input[0].xy__, 2D[0];
>  2: src0.w = temp[0]
>     MAD temp[0].x, src0.w__, src0.111, src0.000
>     MAD color[0].w, src0.1, src0.1, src0.0
>  3: src0.xyz = temp[0], src1.xyz = const[4]
>     MAD temp[0].xy, src0.xy_, src1.xx_, -src0.11_
>  4: 
>     MAD temp[0].z, src0.__0, src0.111, src0.000
>  5: src0.xyz = temp[0]
>     DP3 temp[1].x, src0.xyz, src0.xyz
>  6: src0.xyz = temp[1]
>     MAD temp[1].x, src0.1__, src0.111, -src0.x__
>  7: src0.xyz = temp[1]
>     REPL_ALPHA temp[1].x
>     RSQ, |src0.x|
>  8: src0.xyz = temp[1]
>     REPL_ALPHA temp[0].z
>     RCP, src0.x
>  9: src0.xyz = input[2], src1.xyz = temp[0]
>     DP3 temp[2].x, src0.xyz, src1.xyz
> 10: src0.xyz = input[3], src1.xyz = temp[0]
>     DP3 temp[2].y, src0.xyz, src1.xyz
> 11: src0.xyz = input[4], src1.xyz = temp[0]
>     DP3 temp[2].z, src0.xyz, src1.xyz
> 12: src0.xyz = input[1], src1.xyz = temp[2]
>     DP3 temp[3].xyz, src0.xyz, src1.xyz
> 13: src0.xyz = temp[3]
>     MAD temp[3].xyz, src0.xyz, src0.xyz, src0.000
> 14: src0.xyz = temp[3], src1.xyz = const[5]
>     MAD_SAT temp[3].xyz, src0.xyz, src1.zzz, src0.000
> 15: src0.xyz = temp[3], src1.xyz = const[3]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src0.000
> 16: src0.xyz = temp[3]
>     MAD color[0].xyz, src0.xyz, src0.111, src0.000
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[0].yw, temp[0].xy__, 2D[0];
>  2: src0.w = temp[0]
>     MAD temp[0].x, src0.w__, src0.111, src0.000
>     MAD color[0].w, src0.1, src0.1, src0.0
>  3: src0.xyz = temp[0], src1.xyz = const[4]
>     MAD temp[0].xy, src0.xy_, src1.xx_, -src0.11_
>  4: 
>     MAD temp[0].z, src0.__0, src0.111, src0.000
>  5: src0.xyz = temp[0]
>     DP3 temp[5].x, src0.xyz, src0.xyz
>  6: src0.xyz = temp[5]
>     MAD temp[5].x, src0.1__, src0.111, -src0.x__
>  7: src0.xyz = temp[5]
>     REPL_ALPHA temp[5].x
>     RSQ, |src0.x|
>  8: src0.xyz = temp[5]
>     REPL_ALPHA temp[0].z
>     RCP, src0.x
>  9: src0.xyz = temp[2], src1.xyz = temp[0]
>     DP3 temp[2].x, src0.xyz, src1.xyz
> 10: src0.xyz = temp[3], src1.xyz = temp[0]
>     DP3 temp[2].y, src0.xyz, src1.xyz
> 11: src0.xyz = temp[4], src1.xyz = temp[0]
>     DP3 temp[2].z, src0.xyz, src1.xyz
> 12: src0.xyz = temp[1], src1.xyz = temp[2]
>     DP3 temp[0].xyz, src0.xyz, src1.xyz
> 13: src0.xyz = temp[0]
>     MAD temp[0].xyz, src0.xyz, src0.xyz, src0.000
> 14: src0.xyz = temp[0], src1.xyz = const[5]
>     MAD_SAT temp[0].xyz, src0.xyz, src1.zzz, src0.000
> 15: src0.xyz = temp[0], src1.xyz = const[3]
>     MAD temp[0].xyz, src0.xyz, src1.xyz, src0.000
> 16: src0.xyz = temp[0]
>     MAD color[0].xyz, src0.xyz, src0.111, src0.000
>pc=8*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 14, tex_end: 0  (code_addr: 00400380)
>  TEX:
>    TEX t0, t0, texture[0]   (00008000)
>  0: xyz:  t0  t0  t0 -> t0.x                 (00800000)
>       w:  t0  t0  t0 -> o0.w                 (01000000)
>     xyz:   t0.www      1.0      0.0    op: 00050a8c
>       w:      1.0      1.0      0.0    op: 00040891
>  1: xyz:  t0  c4  t0 -> t0.xy                (01800900)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xyz   c4.xxx     -1.0    op: 000d4280
>       w:     t0.x     t0.x     t0.x    op: 00000000
>  2: xyz:  t0  t0  t0 -> t0.z                 (02000000)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:      0.0      1.0      0.0    op: 00050a94
>       w:     t0.x     t0.x     t0.x    op: 00000000
>  3: xyz:  t0  t0  t0 -> t5.x                 (00940000)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xyz   t0.xyz   t0.xxx    op: 00804000
>       w:     t0.x     t0.x     t0.x    op: 00000000
>  4: xyz:  t5  t0  t0 -> t5.x                 (00940005)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:      1.0      1.0  -t5.xyz    op: 00080a95
>       w:     t5.x     t5.x     t5.x    op: 00000000
>  5: xyz:  t5  t0  t0 -> t5.x                 (00940005)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t5.xxx   t5.xxx   t5.xxx    op: 05004081
>       w:   |t5.x|     t5.x     t5.x    op: 05800040
>  6: xyz:  t5  t0  t0 -> t0.z                 (02000005)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t5.xxx   t5.xxx   t5.xxx    op: 05004081
>       w:     t5.x     t5.x     t5.x    op: 05000000
>  7: xyz:  t2  t0  t0 -> t2.x                 (00880002)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz   t0.xyz   t2.xxx    op: 00804200
>       w:     t2.x     t2.x     t2.x    op: 00000000
>  8: xyz:  t3  t0  t0 -> t2.y                 (01080003)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t3.xyz   t0.xyz   t3.xxx    op: 00804200
>       w:     t3.x     t3.x     t3.x    op: 00000000
>  9: xyz:  t4  t0  t0 -> t2.z                 (02080004)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t4.xyz   t0.xyz   t4.xxx    op: 00804200
>       w:     t4.x     t4.x     t4.x    op: 00000000
> 10: xyz:  t1  t2  t0 -> t0.xyz               (03800081)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   t2.xyz   t1.xxx    op: 00804200
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 11: xyz:  t0  t0  t0 -> t0.xyz               (03800000)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xyz   t0.xyz      0.0    op: 00050000
>       w:     t0.x     t0.x     t0.x    op: 00000000
> 12: xyz:  t0  c5  t0 -> t0.xyz               (03800940)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xyz   c5.zzz      0.0    op: 40050380
>       w:     t0.x     t0.x     t0.x    op: 00000000
> 13: xyz:  t0  c3  t0 -> t0.xyz               (038008c0)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xyz   c3.xyz      0.0    op: 00050200
>       w:     t0.x     t0.x     t0.x    op: 00000000
> 14: xyz:  t0  t0  t0 -> o0.xyz               (1c000000)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xyz      1.0      0.0    op: 00050a80
>       w:     t0.x     t0.x     t0.x    op: 00000000
>r300: Initial fragment program
>FRAG
>DCL IN[0], COLOR, LINEAR
>DCL IN[1], GENERIC[0], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL TEMP[0]
>IMM FLT32 {    0.3330,     1.0000,     0.0000,     0.0000}
>  0: TEX TEMP[0], IN[1], SAMP[0], 2D
>  1: DP3 TEMP[0].w, TEMP[0], IMM[0].xxxx
>  2: MOV TEMP[0].xyz, IMM[0].yyyy
>  3: MUL OUT[0], TEMP[0], IN[0]
>  4: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: DP3 temp[0].w, temp[0], const[0].xxxx;
>  2: MOV temp[0].xyz, const[0].yyyy;
>  3: MUL output[0], temp[0], input[0];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: DP3 temp[0].w, temp[0], const[0].xxxx;
>  2: MOV temp[0].xyz, const[0].yyyy;
>  3: MUL output[0], temp[0], input[0];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: DP3 temp[0].w, temp[0], const[0].xxxx;
>  2: MOV temp[0].xyz, const[0].yyyy;
>  3: MUL output[0], temp[0], input[0];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: DP3 temp[0].w, temp[0], const[0].xxxx;
>  2: MOV temp[0].xyz, const[0].yyyy;
>  3: MUL output[0], temp[0], input[0];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: DP3 temp[0].w, temp[0], const[0].xxxx;
>  2: MOV temp[0].xyz, const[0].yyyy;
>  3: MUL output[0], temp[0], input[0];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: TEX temp[0].xyz, input[1].xy__, 2D[0];
>  1: DP3 temp[0].w, temp[0].xyz_, const[0].xxx_;
>  2: MOV temp[0].xyz, const[0].yyy_;
>  3: MUL output[0], temp[0], input[0];
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: TEX temp[0].xyz, input[1].xy__, 2D[0];
>  1: DP3 temp[0].w, temp[0].xyz_, const[0].xxx_;
>  2: MOV temp[0].xyz, none.111_;
>  3: MUL output[0], temp[0], input[0];
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: TEX temp[0].xyz, input[1].xy__, 2D[0];
>  1: DP3 temp[0].w, temp[0].xyz_, const[0].xxx_;
>  2: MOV temp[0].xyz, none.111_;
>  3: MUL output[0], temp[0], input[0];
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: TEX temp[0].xyz, input[1].xy__, 2D[0];
>  1: src0.xyz = temp[0], src1.xyz = const[0]
>     DP3, src0.xyz, src1.xxx
>     DP3 temp[0].w, src0._, src0._
>  2: 
>     MAD temp[0].xyz, src0.111, src0.111, src0.000
>  3: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = input[0], src1.w = input[0]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[0].xyz, input[1].xy__, 2D[0];
>  2: src0.xyz = temp[0], src1.xyz = const[0]
>     DP3, src0.xyz, src1.xxx
>     DP3 temp[0].w, src0._, src0._
>  3: 
>     MAD temp[0].xyz, src0.111, src0.111, src0.000
>  4: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = input[0], src1.w = input[0]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[1].xyz, temp[1].xy__, 2D[0];
>  2: src0.xyz = temp[1], src1.xyz = const[0]
>     DP3, src0.xyz, src1.xxx
>     DP3 temp[1].w, src0._, src0._
>  3: 
>     MAD temp[1].xyz, src0.111, src0.111, src0.000
>  4: src0.xyz = temp[1], src0.w = temp[1], src1.xyz = temp[0], src1.w = temp[0]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>pc=9*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 2, tex_end: 0  (code_addr: 00400080)
>  TEX:
>    TEX t1, t1, texture[0]   (00008041)
>  0: xyz:  t1  c0  t0 ->                      (00000801)
>       w:  t0  t0  t0 -> t1.w                 (00840000)
>     xyz:   t1.xyz   c0.xxx   t1.xxx    op: 00804280
>       w:      1.0      1.0     t1.x    op: 00800891
>  1: xyz:  t0  t0  t0 -> t1.xyz               (03840000)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:      1.0      1.0      0.0    op: 00050a95
>       w:     t0.x     t0.x     t0.x    op: 00000000
>  2: xyz:  t1  t0  t0 -> o0.xyz               (1c000001)
>       w:  t1  t0  t0 -> o0.w                 (01000001)
>     xyz:   t1.xyz   t0.xyz      0.0    op: 00050200
>       w:     t1.w     t0.w      0.0    op: 00040509
>r300: Initial fragment program
>FRAG
>DCL IN[0], COLOR, LINEAR
>DCL IN[1], GENERIC[0], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL TEMP[0]
>IMM FLT32 {    1.0000,     0.0000,     0.0000,     0.0000}
>  0: MOV TEMP[0].xyz, IMM[0].xxxx
>  1: TEX TEMP[0].w, IN[1], SAMP[0], 2D
>  2: MUL OUT[0], TEMP[0], IN[0]
>  3: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: MOV temp[0].xyz, temp[0].1111;
>  1: TEX temp[0].w, input[1], 2D[0];
>  2: MUL output[0], temp[0], input[0];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: MOV temp[0].xyz, temp[0].1111;
>  1: TEX temp[0].w, input[1], 2D[0];
>  2: MUL output[0], temp[0], input[0];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: MOV temp[0].xyz, temp[0].1111;
>  1: TEX temp[0].w, input[1], 2D[0];
>  2: MUL output[0], temp[0], input[0];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: MOV temp[0].xyz, temp[0].1111;
>  1: TEX temp[0].w, input[1], 2D[0];
>  2: MUL output[0], temp[0], input[0];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: MOV temp[0].xyz, temp[0].1111;
>  1: TEX temp[1], input[1], 2D[0];
>  2: MOV temp[0].w, temp[1];
>  3: MUL output[0], temp[0], input[0];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: MOV temp[0].xyz, temp[0].111_;
>  1: TEX temp[1].w, input[1].xy__, 2D[0];
>  2: MOV temp[0].w, temp[1].___w;
>  3: MUL output[0], temp[0], input[0];
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: MOV temp[0].xyz, temp[0].111_;
>  1: TEX temp[1].w, input[1].xy__, 2D[0];
>  2: MOV temp[0].w, temp[1].___w;
>  3: MUL output[0], temp[0], input[0];
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: MOV temp[0].xyz, temp[0].111_;
>  1: TEX temp[1].w, input[1].xy__, 2D[0];
>  2: MOV temp[0].w, temp[1].___w;
>  3: MUL output[0], temp[0], input[0];
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: 
>     MAD temp[0].xyz, src0.111, src0.111, src0.000
>  1: TEX temp[1].w, input[1].xy__, 2D[0];
>  2: src0.w = temp[1]
>     MAD temp[0].w, src0.w, src0.1, src0.0
>  3: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = input[0], src1.w = input[0]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[1].w, input[1].xy__, 2D[0];
>  2: src0.w = temp[1]
>     MAD temp[0].xyz, src0.111, src0.111, src0.000
>     MAD temp[0].w, src0.w, src0.1, src0.0
>  3: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = input[0], src1.w = input[0]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[1].w, temp[1].xy__, 2D[0];
>  2: src0.w = temp[1]
>     MAD temp[1].xyz, src0.111, src0.111, src0.000
>     MAD temp[1].w, src0.w, src0.1, src0.0
>  3: src0.xyz = temp[1], src0.w = temp[1], src1.xyz = temp[0], src1.w = temp[0]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>pc=10*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 1, tex_end: 0  (code_addr: 00400040)
>  TEX:
>    TEX t1, t1, texture[0]   (00008041)
>  0: xyz:  t0  t0  t0 -> t1.xyz               (03840000)
>       w:  t1  t0  t0 -> t1.w                 (00840001)
>     xyz:      1.0      1.0      0.0    op: 00050a95
>       w:     t1.w      1.0      0.0    op: 00040889
>  1: xyz:  t1  t0  t0 -> o0.xyz               (1c000001)
>       w:  t1  t0  t0 -> o0.w                 (01000001)
>     xyz:   t1.xyz   t0.xyz      0.0    op: 00050200
>       w:     t1.w     t0.w      0.0    op: 00040509
>r300: Initial fragment program
>FRAG
>DCL IN[0], GENERIC[0], LINEAR
>DCL OUT[0], COLOR
>DCL SAMP[0]
>  0: TEX OUT[0], IN[0], SAMP[0], 2D
>  1: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: TEX output[0], input[0], 2D[0];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: TEX output[0], input[0], 2D[0];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: TEX output[0], input[0], 2D[0];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: TEX output[0], input[0], 2D[0];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: TEX temp[1], input[0], 2D[0];
>  1: MOV output[0], temp[1];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: TEX temp[1], input[0].xy__, 2D[0];
>  1: MOV output[0], temp[1];
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: TEX temp[1], input[0].xy__, 2D[0];
>  1: MOV output[0], temp[1];
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: TEX temp[1], input[0].xy__, 2D[0];
>  1: MOV output[0], temp[1];
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: TEX temp[1], input[0].xy__, 2D[0];
>  1: src0.xyz = temp[1], src0.w = temp[1]
>     MAD color[0].xyz, src0.xyz, src0.111, src0.000
>     MAD color[0].w, src0.w, src0.1, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[1], input[0].xy__, 2D[0];
>  2: src0.xyz = temp[1], src0.w = temp[1]
>     MAD color[0].xyz, src0.xyz, src0.111, src0.000
>     MAD color[0].w, src0.w, src0.1, src0.0
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[0], temp[0].xy__, 2D[0];
>  2: src0.xyz = temp[0], src0.w = temp[0]
>     MAD color[0].xyz, src0.xyz, src0.111, src0.000
>     MAD color[0].w, src0.w, src0.1, src0.0
>pc=11*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 0, tex_end: 0  (code_addr: 00400000)
>  TEX:
>    TEX t0, t0, texture[0]   (00008000)
>  0: xyz:  t0  t0  t0 -> o0.xyz               (1c000000)
>       w:  t0  t0  t0 -> o0.w                 (01000000)
>     xyz:   t0.xyz      1.0      0.0    op: 00050a80
>       w:     t0.w      1.0      0.0    op: 00040889
>r300: Initial fragment program
>FRAG
>DCL IN[0], COLOR, PERSPECTIVE
>DCL OUT[0], COLOR
>  0: MOV OUT[0], IN[0]
>  1: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: src0.xyz = input[0], src0.w = input[0]
>     MAD color[0].xyz, src0.xyz, src0.111, src0.000
>     MAD color[0].w, src0.w, src0.1, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: src0.xyz = input[0], src0.w = input[0]
>     MAD color[0].xyz, src0.xyz, src0.111, src0.000
>     MAD color[0].w, src0.w, src0.1, src0.0
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: src0.xyz = temp[0], src0.w = temp[0]
>     MAD color[0].xyz, src0.xyz, src0.111, src0.000
>     MAD color[0].w, src0.w, src0.1, src0.0
>pc=12*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 0, tex_end: 0  (code_addr: 00400000)
>  0: xyz:  t0  t0  t0 -> o0.xyz               (1c000000)
>       w:  t0  t0  t0 -> o0.w                 (01000000)
>     xyz:   t0.xyz      1.0      0.0    op: 00050a80
>       w:     t0.w      1.0      0.0    op: 00040889
>r300: Initial fragment program
>FRAG
>DCL IN[0], GENERIC[0], LINEAR
>DCL OUT[0], COLOR
>DCL SAMP[0]
>  0: TEX OUT[0], IN[0], SAMP[0], 2D
>  1: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: TEX output[0], input[0], 2D[0];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: TEX output[0], input[0], 2D[0];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: TEX output[0], input[0], 2D[0];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: TEX output[0], input[0], 2D[0];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: TEX temp[1], input[0], 2D[0];
>  1: MOV output[0], temp[1];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: TEX temp[1], input[0].xy__, 2D[0];
>  1: MOV output[0], temp[1];
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: TEX temp[1], input[0].xy__, 2D[0];
>  1: MOV output[0], temp[1];
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: TEX temp[1], input[0].xy__, 2D[0];
>  1: MOV output[0], temp[1];
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: TEX temp[1], input[0].xy__, 2D[0];
>  1: src0.xyz = temp[1], src0.w = temp[1]
>     MAD color[0].xyz, src0.xyz, src0.111, src0.000
>     MAD color[0].w, src0.w, src0.1, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[1], input[0].xy__, 2D[0];
>  2: src0.xyz = temp[1], src0.w = temp[1]
>     MAD color[0].xyz, src0.xyz, src0.111, src0.000
>     MAD color[0].w, src0.w, src0.1, src0.0
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[0], temp[0].xy__, 2D[0];
>  2: src0.xyz = temp[0], src0.w = temp[0]
>     MAD color[0].xyz, src0.xyz, src0.111, src0.000
>     MAD color[0].w, src0.w, src0.1, src0.0
>pc=13*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 0, tex_end: 0  (code_addr: 00400000)
>  TEX:
>    TEX t0, t0, texture[0]   (00008000)
>  0: xyz:  t0  t0  t0 -> o0.xyz               (1c000000)
>       w:  t0  t0  t0 -> o0.w                 (01000000)
>     xyz:   t0.xyz      1.0      0.0    op: 00050a80
>       w:     t0.w      1.0      0.0    op: 00040889
>r300: Initial fragment program
>FRAG
>DCL IN[0], GENERIC[0], PERSPECTIVE
>DCL IN[1], GENERIC[1], PERSPECTIVE
>DCL IN[2], GENERIC[2], PERSPECTIVE
>DCL IN[3], GENERIC[3], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL CONST[1]
>DCL TEMP[0..2]
>IMM FLT32 {    0.2500,     0.3300,     1.0000,     2.0000}
>IMM FLT32 {    3.0000,     0.0000,     1.0000,     0.0000}
>  0: TEX TEMP[0], IN[1], SAMP[0], RECT
>  1: TEX TEMP[1], IN[0], SAMP[0], RECT
>  2: ADD TEMP[2], TEMP[1], TEMP[0]
>  3: TEX TEMP[1], IN[2], SAMP[0], RECT
>  4: ADD TEMP[1], TEMP[2], TEMP[1]
>  5: TEX TEMP[0], IN[3], SAMP[0], RECT
>  6: ADD TEMP[0], TEMP[1], TEMP[0]
>  7: MOV TEMP[1].x, IMM[0].zzzz
>  8: ADD TEMP[1].y, -CONST[1].zzzz, TEMP[1].xxxx
>  9: MUL TEMP[0], TEMP[0], IMM[0].xxxx
> 10: DP3 TEMP[1].x, TEMP[0], IMM[0].yyyy
> 11: RCP TEMP[1].y, TEMP[1].yyyy
> 12: ADD TEMP[1].x, TEMP[1], -CONST[1].zzzz
> 13: MUL_SAT TEMP[1].x, TEMP[1], TEMP[1].yyyy
> 14: MUL TEMP[1].y, -TEMP[1].xxxx, IMM[0].wwww
> 15: ADD TEMP[1].y, TEMP[1], IMM[1].xxxx
> 16: MUL TEMP[1].x, TEMP[1], TEMP[1]
> 17: MAD TEMP[1].x, TEMP[1], TEMP[1].yyyy, -IMM[0].zzzz
> 18: MUL TEMP[1].x, TEMP[1], CONST[1].wwww
> 19: ADD TEMP[1].x, TEMP[1], IMM[0].zzzz
> 20: MOV TEMP[1].w, IMM[0].zzzz
> 21: MUL TEMP[1].xyz, TEMP[1].xxxx, TEMP[0].wwww
> 22: MUL OUT[0], TEMP[0], TEMP[1]
> 23: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: TEX temp[0], input[1], RECT[0];
>  1: TEX temp[1], input[0], RECT[0];
>  2: ADD temp[2], temp[1], temp[0];
>  3: TEX temp[1], input[2], RECT[0];
>  4: ADD temp[1], temp[2], temp[1];
>  5: TEX temp[0], input[3], RECT[0];
>  6: ADD temp[0], temp[1], temp[0];
>  7: MOV temp[1].x, const[2].zzzz;
>  8: ADD temp[1].y, -const[1].zzzz, temp[1].xxxx;
>  9: MUL temp[0], temp[0], const[2].xxxx;
> 10: DP3 temp[1].x, temp[0], const[2].yyyy;
> 11: RCP temp[1].y, temp[1].yyyy;
> 12: ADD temp[1].x, temp[1], -const[1].zzzz;
> 13: MUL_SAT temp[1].x, temp[1], temp[1].yyyy;
> 14: MUL temp[1].y, -temp[1].xxxx, const[2].wwww;
> 15: ADD temp[1].y, temp[1], const[3].xxxx;
> 16: MUL temp[1].x, temp[1], temp[1];
> 17: MAD temp[1].x, temp[1], temp[1].yyyy, -const[2].zzzz;
> 18: MUL temp[1].x, temp[1], const[1].wwww;
> 19: ADD temp[1].x, temp[1], const[2].zzzz;
> 20: MOV temp[1].w, const[2].zzzz;
> 21: MUL temp[1].xyz, temp[1].xxxx, temp[0].wwww;
> 22: MUL output[0], temp[0], temp[1];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: TEX temp[0], input[1], RECT[0];
>  1: TEX temp[1], input[0], RECT[0];
>  2: ADD temp[2], temp[1], temp[0];
>  3: TEX temp[1], input[2], RECT[0];
>  4: ADD temp[1], temp[2], temp[1];
>  5: TEX temp[0], input[3], RECT[0];
>  6: ADD temp[0], temp[1], temp[0];
>  7: MOV temp[1].x, const[2].zzzz;
>  8: ADD temp[1].y, -const[1].zzzz, temp[1].xxxx;
>  9: MUL temp[0], temp[0], const[2].xxxx;
> 10: DP3 temp[1].x, temp[0], const[2].yyyy;
> 11: RCP temp[1].y, temp[1].yyyy;
> 12: ADD temp[1].x, temp[1], -const[1].zzzz;
> 13: MUL_SAT temp[1].x, temp[1], temp[1].yyyy;
> 14: MUL temp[1].y, -temp[1].xxxx, const[2].wwww;
> 15: ADD temp[1].y, temp[1], const[3].xxxx;
> 16: MUL temp[1].x, temp[1], temp[1];
> 17: MAD temp[1].x, temp[1], temp[1].yyyy, -const[2].zzzz;
> 18: MUL temp[1].x, temp[1], const[1].wwww;
> 19: ADD temp[1].x, temp[1], const[2].zzzz;
> 20: MOV temp[1].w, const[2].zzzz;
> 21: MUL temp[1].xyz, temp[1].xxxx, temp[0].wwww;
> 22: MUL output[0], temp[0], temp[1];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: TEX temp[0], input[1], RECT[0];
>  1: TEX temp[1], input[0], RECT[0];
>  2: ADD temp[2], temp[1], temp[0];
>  3: TEX temp[1], input[2], RECT[0];
>  4: ADD temp[1], temp[2], temp[1];
>  5: TEX temp[0], input[3], RECT[0];
>  6: ADD temp[0], temp[1], temp[0];
>  7: MOV temp[1].x, const[2].zzzz;
>  8: ADD temp[1].y, -const[1].zzzz, temp[1].xxxx;
>  9: MUL temp[0], temp[0], const[2].xxxx;
> 10: DP3 temp[1].x, temp[0], const[2].yyyy;
> 11: RCP temp[1].y, temp[1].yyyy;
> 12: ADD temp[1].x, temp[1], -const[1].zzzz;
> 13: MUL_SAT temp[1].x, temp[1], temp[1].yyyy;
> 14: MUL temp[1].y, -temp[1].xxxx, const[2].wwww;
> 15: ADD temp[1].y, temp[1], const[3].xxxx;
> 16: MUL temp[1].x, temp[1], temp[1];
> 17: MAD temp[1].x, temp[1], temp[1].yyyy, -const[2].zzzz;
> 18: MUL temp[1].x, temp[1], const[1].wwww;
> 19: ADD temp[1].x, temp[1], const[2].zzzz;
> 20: MOV temp[1].w, const[2].zzzz;
> 21: MUL temp[1].xyz, temp[1].xxxx, temp[0].wwww;
> 22: MUL output[0], temp[0], temp[1];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: TEX temp[0], input[1], RECT[0];
>  1: TEX temp[1], input[0], RECT[0];
>  2: ADD temp[2], temp[1], temp[0];
>  3: TEX temp[1], input[2], RECT[0];
>  4: ADD temp[1], temp[2], temp[1];
>  5: TEX temp[0], input[3], RECT[0];
>  6: ADD temp[0], temp[1], temp[0];
>  7: MOV temp[1].x, const[2].zzzz;
>  8: ADD temp[1].y, -const[1].zzzz, temp[1].xxxx;
>  9: MUL temp[0], temp[0], const[2].xxxx;
> 10: DP3 temp[1].x, temp[0], const[2].yyyy;
> 11: RCP temp[1].y, temp[1].yyyy;
> 12: ADD temp[1].x, temp[1], -const[1].zzzz;
> 13: MUL_SAT temp[1].x, temp[1], temp[1].yyyy;
> 14: MUL temp[1].y, -temp[1].xxxx, const[2].wwww;
> 15: ADD temp[1].y, temp[1], const[3].xxxx;
> 16: MUL temp[1].x, temp[1], temp[1];
> 17: MAD temp[1].x, temp[1], temp[1].yyyy, -const[2].zzzz;
> 18: MUL temp[1].x, temp[1], const[1].wwww;
> 19: ADD temp[1].x, temp[1], const[2].zzzz;
> 20: MOV temp[1].w, const[2].zzzz;
> 21: MUL temp[1].xyz, temp[1].xxxx, temp[0].wwww;
> 22: MUL output[0], temp[0], temp[1];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: MUL temp[3], input[1], const[4];
>  1: TEX temp[0], temp[3], 2D[0];
>  2: MUL temp[4], input[0], const[4];
>  3: TEX temp[1], temp[4], 2D[0];
>  4: ADD temp[2], temp[1], temp[0];
>  5: MUL temp[5], input[2], const[4];
>  6: TEX temp[1], temp[5], 2D[0];
>  7: ADD temp[1], temp[2], temp[1];
>  8: MUL temp[6], input[3], const[4];
>  9: TEX temp[0], temp[6], 2D[0];
> 10: ADD temp[0], temp[1], temp[0];
> 11: MOV temp[1].x, const[2].zzzz;
> 12: ADD temp[1].y, -const[1].zzzz, temp[1].xxxx;
> 13: MUL temp[0], temp[0], const[2].xxxx;
> 14: DP3 temp[1].x, temp[0], const[2].yyyy;
> 15: RCP temp[1].y, temp[1].yyyy;
> 16: ADD temp[1].x, temp[1], -const[1].zzzz;
> 17: MUL_SAT temp[1].x, temp[1], temp[1].yyyy;
> 18: MUL temp[1].y, -temp[1].xxxx, const[2].wwww;
> 19: ADD temp[1].y, temp[1], const[3].xxxx;
> 20: MUL temp[1].x, temp[1], temp[1];
> 21: MAD temp[1].x, temp[1], temp[1].yyyy, -const[2].zzzz;
> 22: MUL temp[1].x, temp[1], const[1].wwww;
> 23: ADD temp[1].x, temp[1], const[2].zzzz;
> 24: MOV temp[1].w, const[2].zzzz;
> 25: MUL temp[1].xyz, temp[1].xxxx, temp[0].wwww;
> 26: MUL output[0], temp[0], temp[1];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: MUL temp[3].xy, input[1].xy__, const[4].xy__;
>  1: TEX temp[0], temp[3].xy__, 2D[0];
>  2: MUL temp[4].xy, input[0].xy__, const[4].xy__;
>  3: TEX temp[1], temp[4].xy__, 2D[0];
>  4: ADD temp[2], temp[1], temp[0];
>  5: MUL temp[5].xy, input[2].xy__, const[4].xy__;
>  6: TEX temp[1], temp[5].xy__, 2D[0];
>  7: ADD temp[1], temp[2], temp[1];
>  8: MUL temp[6].xy, input[3].xy__, const[4].xy__;
>  9: TEX temp[0], temp[6].xy__, 2D[0];
> 10: ADD temp[0], temp[1], temp[0];
> 11: MOV temp[1].x, const[2].z___;
> 12: ADD temp[1].y, -const[1]._z__, temp[1]._x__;
> 13: MUL temp[0], temp[0], const[2].xxxx;
> 14: DP3 temp[1].x, temp[0].xyz_, const[2].yyy_;
> 15: RCP temp[1].y, temp[1].y___;
> 16: ADD temp[1].x, temp[1].x___, -const[1].z___;
> 17: MUL_SAT temp[1].x, temp[1].x___, temp[1].y___;
> 18: MUL temp[1].y, -temp[1]._x__, const[2]._w__;
> 19: ADD temp[1].y, temp[1]._y__, const[3]._x__;
> 20: MUL temp[1].x, temp[1].x___, temp[1].x___;
> 21: MAD temp[1].x, temp[1].x___, temp[1].y___, -const[2].z___;
> 22: MUL temp[1].x, temp[1].x___, const[1].w___;
> 23: ADD temp[1].x, temp[1].x___, const[2].z___;
> 24: MOV temp[1].w, const[2].___z;
> 25: MUL temp[1].xyz, temp[1].xxx_, temp[0].www_;
> 26: MUL output[0], temp[0], temp[1];
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: MUL temp[3].xy, input[1].xy__, const[4].xy__;
>  1: TEX temp[0], temp[3].xy__, 2D[0];
>  2: MUL temp[4].xy, input[0].xy__, const[4].xy__;
>  3: TEX temp[1], temp[4].xy__, 2D[0];
>  4: ADD temp[2], temp[1], temp[0];
>  5: MUL temp[5].xy, input[2].xy__, const[4].xy__;
>  6: TEX temp[1], temp[5].xy__, 2D[0];
>  7: ADD temp[1], temp[2], temp[1];
>  8: MUL temp[6].xy, input[3].xy__, const[4].xy__;
>  9: TEX temp[0], temp[6].xy__, 2D[0];
> 10: ADD temp[0], temp[1], temp[0];
> 11: ADD temp[1].y, -const[1]._z__, none._1__;
> 12: MUL temp[0], temp[0], const[2].xxxx;
> 13: DP3 temp[1].x, temp[0].xyz_, const[2].yyy_;
> 14: RCP temp[1].y, temp[1].y___;
> 15: ADD temp[1].x, temp[1].x___, -const[1].z___;
> 16: MUL_SAT temp[1].x, temp[1].x___, temp[1].y___;
> 17: MUL temp[1].y, -temp[1]._x__, const[2]._w__;
> 18: ADD temp[1].y, temp[1]._y__, const[3]._x__;
> 19: MUL temp[1].x, temp[1].x___, temp[1].x___;
> 20: MAD temp[1].x, temp[1].x___, temp[1].y___, -none.1___;
> 21: MUL temp[1].x, temp[1].x___, const[1].w___;
> 22: ADD temp[1].x, temp[1].x___, none.1___;
> 23: MOV temp[1].w, none.___1;
> 24: MUL temp[1].xyz, temp[1].xxx_, temp[0].www_;
> 25: MUL output[0], temp[0], temp[1];
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: MUL temp[3].xy, input[1].xy__, const[4].xy__;
>  1: TEX temp[0], temp[3].xy__, 2D[0];
>  2: MUL temp[4].xy, input[0].xy__, const[4].xy__;
>  3: TEX temp[1], temp[4].xy__, 2D[0];
>  4: ADD temp[2], temp[1], temp[0];
>  5: MUL temp[5].xy, input[2].xy__, const[4].xy__;
>  6: TEX temp[1], temp[5].xy__, 2D[0];
>  7: ADD temp[1], temp[2], temp[1];
>  8: MUL temp[6].xy, input[3].xy__, const[4].xy__;
>  9: TEX temp[0], temp[6].xy__, 2D[0];
> 10: ADD temp[0], temp[1], temp[0];
> 11: ADD temp[1].y, -const[1]._z__, none._1__;
> 12: MUL temp[0], temp[0], const[2].xxxx;
> 13: DP3 temp[1].x, temp[0].xyz_, const[2].yyy_;
> 14: RCP temp[1].y, temp[1].y___;
> 15: ADD temp[1].x, temp[1].x___, -const[1].z___;
> 16: MUL_SAT temp[1].x, temp[1].x___, temp[1].y___;
> 17: MUL temp[1].y, -temp[1]._x__, const[2]._w__;
> 18: ADD temp[1].y, temp[1]._y__, const[3]._x__;
> 19: MUL temp[1].x, temp[1].x___, temp[1].x___;
> 20: MAD temp[1].x, temp[1].x___, temp[1].y___, -none.1___;
> 21: MUL temp[1].x, temp[1].x___, const[1].w___;
> 22: ADD temp[1].x, temp[1].x___, none.1___;
> 23: MOV temp[1].w, none.___1;
> 24: MUL temp[1].xyz, temp[1].xxx_, temp[0].www_;
> 25: MUL output[0], temp[0], temp[1];
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: src0.xyz = input[1], src1.xyz = const[4]
>     MAD temp[3].xy, src0.xy_, src1.xy_, src0.000
>  1: TEX temp[0], temp[3].xy__, 2D[0];
>  2: src0.xyz = input[0], src1.xyz = const[4]
>     MAD temp[4].xy, src0.xy_, src1.xy_, src0.000
>  3: TEX temp[1], temp[4].xy__, 2D[0];
>  4: src0.xyz = temp[1], src0.w = temp[1], src1.xyz = temp[0], src1.w = temp[0]
>     MAD temp[2].xyz, src0.xyz, src0.111, src1.xyz
>     MAD temp[2].w, src0.w, src0.1, src1.w
>  5: src0.xyz = input[2], src1.xyz = const[4]
>     MAD temp[5].xy, src0.xy_, src1.xy_, src0.000
>  6: TEX temp[1], temp[5].xy__, 2D[0];
>  7: src0.xyz = temp[2], src0.w = temp[2], src1.xyz = temp[1], src1.w = temp[1]
>     MAD temp[1].xyz, src0.xyz, src0.111, src1.xyz
>     MAD temp[1].w, src0.w, src0.1, src1.w
>  8: src0.xyz = input[3], src1.xyz = const[4]
>     MAD temp[6].xy, src0.xy_, src1.xy_, src0.000
>  9: TEX temp[0], temp[6].xy__, 2D[0];
> 10: src0.xyz = temp[1], src0.w = temp[1], src1.xyz = temp[0], src1.w = temp[0]
>     MAD temp[0].xyz, src0.xyz, src0.111, src1.xyz
>     MAD temp[0].w, src0.w, src0.1, src1.w
> 11: src0.xyz = const[1]
>     MAD temp[1].y, -src0._z_, src0.111, src0._1_
> 12: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[2]
>     MAD temp[0].xyz, src0.xyz, src1.xxx, src0.000
>     MAD temp[0].w, src0.w, src1.x, src0.0
> 13: src0.xyz = temp[0], src1.xyz = const[2]
>     DP3 temp[1].x, src0.xyz, src1.yyy
> 14: src0.xyz = temp[1]
>     REPL_ALPHA temp[1].y
>     RCP, src0.y
> 15: src0.xyz = temp[1], src1.xyz = const[1]
>     MAD temp[1].x, src0.x__, src0.111, -src1.z__
> 16: src0.xyz = temp[1]
>     MAD_SAT temp[1].x, src0.x__, src0.y__, src0.000
> 17: src0.xyz = temp[1], src0.w = const[2]
>     MAD temp[1].y, -src0._x_, src0._w_, src0.000
> 18: src0.xyz = temp[1], src1.xyz = const[3]
>     MAD temp[1].y, src0._y_, src0.111, src1._x_
> 19: src0.xyz = temp[1]
>     MAD temp[1].x, src0.x__, src0.x__, src0.000
> 20: src0.xyz = temp[1]
>     MAD temp[1].x, src0.x__, src0.y__, -src0.1__
> 21: src0.xyz = temp[1], src0.w = const[1]
>     MAD temp[1].x, src0.x__, src0.w__, src0.000
> 22: src0.xyz = temp[1]
>     MAD temp[1].x, src0.x__, src0.111, src0.1__
> 23: 
>     MAD temp[1].w, src0.1, src0.1, src0.0
> 24: src0.xyz = temp[1], src0.w = temp[0]
>     MAD temp[1].xyz, src0.xxx, src0.www, src0.000
> 25: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = temp[1], src1.w = temp[1]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: src0.xyz = input[3], src1.xyz = const[4]
>     MAD temp[6].xy, src0.xy_, src1.xy_, src0.000
>  1: src0.xyz = input[2], src1.xyz = const[4]
>     MAD temp[5].xy, src0.xy_, src1.xy_, src0.000
>  2: src0.xyz = input[0], src1.xyz = const[4]
>     MAD temp[4].xy, src0.xy_, src1.xy_, src0.000
>  3: src0.xyz = input[1], src1.xyz = const[4]
>     MAD temp[3].xy, src0.xy_, src1.xy_, src0.000
>  4: BEGIN_TEX;
>  5: TEX temp[0], temp[3].xy__, 2D[0];
>  6: TEX temp[1], temp[4].xy__, 2D[0];
>  7: src0.xyz = temp[1], src0.w = temp[1], src1.xyz = temp[0], src1.w = temp[0]
>     MAD temp[2].xyz, src0.xyz, src0.111, src1.xyz
>     MAD temp[2].w, src0.w, src0.1, src1.w
>  8: BEGIN_TEX;
>  9: TEX temp[0], temp[6].xy__, 2D[0];
> 10: TEX temp[1], temp[5].xy__, 2D[0];
> 11: src0.xyz = temp[2], src0.w = temp[2], src1.xyz = temp[1], src1.w = temp[1]
>     MAD temp[1].xyz, src0.xyz, src0.111, src1.xyz
>     MAD temp[1].w, src0.w, src0.1, src1.w
> 12: src0.xyz = temp[1], src0.w = temp[1], src1.xyz = temp[0], src1.w = temp[0]
>     MAD temp[0].xyz, src0.xyz, src0.111, src1.xyz
>     MAD temp[0].w, src0.w, src0.1, src1.w
> 13: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[2]
>     MAD temp[0].xyz, src0.xyz, src1.xxx, src0.000
>     MAD temp[0].w, src0.w, src1.x, src0.0
> 14: src0.xyz = temp[0], src1.xyz = const[2]
>     DP3 temp[1].x, src0.xyz, src1.yyy
>     MAD temp[1].w, src0.1, src0.1, src0.0
> 15: src0.xyz = temp[1], src1.xyz = const[1]
>     MAD temp[1].x, src0.x__, src0.111, -src1.z__
> 16: src0.xyz = const[1]
>     MAD temp[1].y, -src0._z_, src0.111, src0._1_
> 17: src0.xyz = temp[1]
>     REPL_ALPHA temp[1].y
>     RCP, src0.y
> 18: src0.xyz = temp[1]
>     MAD_SAT temp[1].x, src0.x__, src0.y__, src0.000
> 19: src0.xyz = temp[1], src0.w = const[2]
>     MAD temp[1].y, -src0._x_, src0._w_, src0.000
> 20: src0.xyz = temp[1], src1.xyz = const[3]
>     MAD temp[1].y, src0._y_, src0.111, src1._x_
> 21: src0.xyz = temp[1]
>     MAD temp[1].x, src0.x__, src0.x__, src0.000
> 22: src0.xyz = temp[1]
>     MAD temp[1].x, src0.x__, src0.y__, -src0.1__
> 23: src0.xyz = temp[1], src0.w = const[1]
>     MAD temp[1].x, src0.x__, src0.w__, src0.000
> 24: src0.xyz = temp[1]
>     MAD temp[1].x, src0.x__, src0.111, src0.1__
> 25: src0.xyz = temp[1], src0.w = temp[0]
>     MAD temp[1].xyz, src0.xxx, src0.www, src0.000
> 26: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = temp[1], src1.w = temp[1]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: src0.xyz = temp[3], src1.xyz = const[4]
>     MAD temp[4].xy, src0.xy_, src1.xy_, src0.000
>  1: src0.xyz = temp[2], src1.xyz = const[4]
>     MAD temp[3].xy, src0.xy_, src1.xy_, src0.000
>  2: src0.xyz = temp[0], src1.xyz = const[4]
>     MAD temp[2].xy, src0.xy_, src1.xy_, src0.000
>  3: src0.xyz = temp[1], src1.xyz = const[4]
>     MAD temp[0].xy, src0.xy_, src1.xy_, src0.000
>  4: BEGIN_TEX;
>  5: TEX temp[0], temp[0].xy__, 2D[0];
>  6: TEX temp[1], temp[2].xy__, 2D[0];
>  7: src0.xyz = temp[1], src0.w = temp[1], src1.xyz = temp[0], src1.w = temp[0]
>     MAD temp[2].xyz, src0.xyz, src0.111, src1.xyz
>     MAD temp[2].w, src0.w, src0.1, src1.w
>  8: BEGIN_TEX;
>  9: TEX temp[0], temp[4].xy__, 2D[0];
> 10: TEX temp[1], temp[3].xy__, 2D[0];
> 11: src0.xyz = temp[2], src0.w = temp[2], src1.xyz = temp[1], src1.w = temp[1]
>     MAD temp[1].xyz, src0.xyz, src0.111, src1.xyz
>     MAD temp[1].w, src0.w, src0.1, src1.w
> 12: src0.xyz = temp[1], src0.w = temp[1], src1.xyz = temp[0], src1.w = temp[0]
>     MAD temp[0].xyz, src0.xyz, src0.111, src1.xyz
>     MAD temp[0].w, src0.w, src0.1, src1.w
> 13: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[2]
>     MAD temp[0].xyz, src0.xyz, src1.xxx, src0.000
>     MAD temp[0].w, src0.w, src1.x, src0.0
> 14: src0.xyz = temp[0], src1.xyz = const[2]
>     DP3 temp[1].x, src0.xyz, src1.yyy
>     MAD temp[1].w, src0.1, src0.1, src0.0
> 15: src0.xyz = temp[1], src1.xyz = const[1]
>     MAD temp[1].x, src0.x__, src0.111, -src1.z__
> 16: src0.xyz = const[1]
>     MAD temp[1].y, -src0._z_, src0.111, src0._1_
> 17: src0.xyz = temp[1]
>     REPL_ALPHA temp[1].y
>     RCP, src0.y
> 18: src0.xyz = temp[1]
>     MAD_SAT temp[1].x, src0.x__, src0.y__, src0.000
> 19: src0.xyz = temp[1], src0.w = const[2]
>     MAD temp[1].y, -src0._x_, src0._w_, src0.000
> 20: src0.xyz = temp[1], src1.xyz = const[3]
>     MAD temp[1].y, src0._y_, src0.111, src1._x_
> 21: src0.xyz = temp[1]
>     MAD temp[1].x, src0.x__, src0.x__, src0.000
> 22: src0.xyz = temp[1]
>     MAD temp[1].x, src0.x__, src0.y__, -src0.1__
> 23: src0.xyz = temp[1], src0.w = const[1]
>     MAD temp[1].x, src0.x__, src0.w__, src0.000
> 24: src0.xyz = temp[1]
>     MAD temp[1].x, src0.x__, src0.111, src0.1__
> 25: src0.xyz = temp[1], src0.w = temp[0]
>     MAD temp[1].xyz, src0.xxx, src0.www, src0.000
> 26: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = temp[1], src1.w = temp[1]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>pc=14*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 3, tex_end: 0  (code_addr: 000000c0)
>  0: xyz:  t3  c4  t0 -> t4.xy                (01900903)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t3.xyz   c4.xyz      0.0    op: 00050200
>       w:     t3.x     t3.x     t3.x    op: 00000000
>  1: xyz:  t2  c4  t0 -> t3.xy                (018c0902)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz   c4.xyz      0.0    op: 00050200
>       w:     t2.x     t2.x     t2.x    op: 00000000
>  2: xyz:  t0  c4  t0 -> t2.xy                (01880900)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xyz   c4.xyz      0.0    op: 00050200
>       w:     t0.x     t0.x     t0.x    op: 00000000
>  3: xyz:  t1  c4  t0 -> t0.xy                (01800901)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   c4.xyz      0.0    op: 00050200
>       w:     t1.x     t1.x     t1.x    op: 00000000
>NODE 1: alu_offset: 4, tex_offset: 0, alu_end: 0, tex_end: 1  (code_addr: 00020004)
>  TEX:
>    TEX t0, t0, texture[0]   (00008000)
>    TEX t1, t2, texture[0]   (00008042)
>  4: xyz:  t1  t0  t0 -> t2.xyz               (03880001)
>       w:  t1  t0  t0 -> t2.w                 (00880001)
>     xyz:   t1.xyz      1.0   t0.xyz    op: 00010a80
>       w:     t1.w      1.0     t0.w    op: 00028889
>NODE 2: alu_offset: 5, tex_offset: 2, alu_end: 15, tex_end: 1  (code_addr: 004223c5)
>  TEX:
>    TEX t0, t4, texture[0]   (00008004)
>    TEX t1, t3, texture[0]   (00008043)
>  5: xyz:  t2  t1  t0 -> t1.xyz               (03840042)
>       w:  t2  t1  t0 -> t1.w                 (00840042)
>     xyz:   t2.xyz      1.0   t1.xyz    op: 00010a80
>       w:     t2.w      1.0     t1.w    op: 00028889
>  6: xyz:  t1  t0  t0 -> t0.xyz               (03800001)
>       w:  t1  t0  t0 -> t0.w                 (00800001)
>     xyz:   t1.xyz      1.0   t0.xyz    op: 00010a80
>       w:     t1.w      1.0     t0.w    op: 00028889
>  7: xyz:  t0  c2  t0 -> t0.xyz               (03800880)
>       w:  t0  t0  t0 -> t0.w                 (00800000)
>     xyz:   t0.xyz   c2.xxx      0.0    op: 00050280
>       w:     t0.w     c2.x      0.0    op: 00040189
>  8: xyz:  t0  c2  t0 -> t1.x                 (00840880)
>       w:  t0  t0  t0 -> t1.w                 (00840000)
>     xyz:   t0.xyz   c2.yyy   t0.xxx    op: 00804300
>       w:      1.0      1.0      0.0    op: 00040891
>  9: xyz:  t1  c1  t0 -> t1.x                 (00840841)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz      1.0  -c1.zzz    op: 0009ca80
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 10: xyz:  c1  t0  t0 -> t1.y                 (01040021)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:  -c1.zzz      1.0      1.0    op: 00054aa3
>       w:     c1.x     c1.x     c1.x    op: 00000000
> 11: xyz:  t1  t0  t0 -> t1.y                 (01040001)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xxx   t1.xxx   t1.xxx    op: 05004081
>       w:     t1.y     t1.x     t1.x    op: 05000001
> 12: xyz:  t1  t0  t0 -> t1.x                 (00840001)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   t1.yyy      0.0    op: 40050100
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 13: xyz:  t1  t0  t0 -> t1.y                 (01040001)
>       w:  c2  t0  t0 ->                      (00000022)
>     xyz:  -t1.xxx   c2.www      0.0    op: 00050621
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 14: xyz:  t1  c3  t0 -> t1.y                 (010408c1)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz      1.0   c3.xxx    op: 00014a80
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 15: xyz:  t1  t0  t0 -> t1.x                 (00840001)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   t1.xyz      0.0    op: 00050000
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 16: xyz:  t1  t0  t0 -> t1.x                 (00840001)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   t1.yyy     -1.0    op: 000d4100
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 17: xyz:  t1  t0  t0 -> t1.x                 (00840001)
>       w:  c1  t0  t0 ->                      (00000021)
>     xyz:   t1.xyz   c1.www      0.0    op: 00050600
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 18: xyz:  t1  t0  t0 -> t1.x                 (00840001)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz      1.0      1.0    op: 00054a80
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 19: xyz:  t1  t0  t0 -> t1.xyz               (03840001)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xxx   t0.www      0.0    op: 00050601
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 20: xyz:  t0  t1  t0 -> o0.xyz               (1c000040)
>       w:  t0  t1  t0 -> o0.w                 (01000040)
>     xyz:   t0.xyz   t1.xyz      0.0    op: 00050200
>       w:     t0.w     t1.w      0.0    op: 00040509
>r300: Initial fragment program
>FRAG
>DCL IN[0], GENERIC[0], PERSPECTIVE
>DCL IN[1], GENERIC[1], PERSPECTIVE
>DCL IN[2], GENERIC[2], PERSPECTIVE
>DCL IN[3], GENERIC[3], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL CONST[1]
>DCL TEMP[0..2]
>IMM FLT32 {    0.2500,     0.3300,     1.0000,     2.0000}
>IMM FLT32 {    3.0000,     0.0000,     1.0000,     0.0000}
>  0: TEX TEMP[0], IN[1], SAMP[0], RECT
>  1: TEX TEMP[1], IN[0], SAMP[0], RECT
>  2: ADD TEMP[2], TEMP[1], TEMP[0]
>  3: TEX TEMP[1], IN[2], SAMP[0], RECT
>  4: ADD TEMP[1], TEMP[2], TEMP[1]
>  5: TEX TEMP[0], IN[3], SAMP[0], RECT
>  6: ADD TEMP[0], TEMP[1], TEMP[0]
>  7: MOV TEMP[1].x, IMM[0].zzzz
>  8: ADD TEMP[1].y, -CONST[1].zzzz, TEMP[1].xxxx
>  9: MUL TEMP[0], TEMP[0], IMM[0].xxxx
> 10: DP3 TEMP[1].x, TEMP[0], IMM[0].yyyy
> 11: RCP TEMP[1].y, TEMP[1].yyyy
> 12: ADD TEMP[1].x, TEMP[1], -CONST[1].zzzz
> 13: MUL_SAT TEMP[1].x, TEMP[1], TEMP[1].yyyy
> 14: MUL TEMP[1].y, -TEMP[1].xxxx, IMM[0].wwww
> 15: ADD TEMP[1].y, TEMP[1], IMM[1].xxxx
> 16: MUL TEMP[1].x, TEMP[1], TEMP[1]
> 17: MAD TEMP[1].x, TEMP[1], TEMP[1].yyyy, -IMM[0].zzzz
> 18: MUL TEMP[1].x, TEMP[1], CONST[1].wwww
> 19: ADD TEMP[1].x, TEMP[1], IMM[0].zzzz
> 20: MOV TEMP[1].w, IMM[0].zzzz
> 21: MUL TEMP[1].xyz, TEMP[1].xxxx, TEMP[0].wwww
> 22: MUL OUT[0], TEMP[0], TEMP[1]
> 23: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: TEX temp[0], input[1], RECT[0];
>  1: TEX temp[1], input[0], RECT[0];
>  2: ADD temp[2], temp[1], temp[0];
>  3: TEX temp[1], input[2], RECT[0];
>  4: ADD temp[1], temp[2], temp[1];
>  5: TEX temp[0], input[3], RECT[0];
>  6: ADD temp[0], temp[1], temp[0];
>  7: MOV temp[1].x, const[2].zzzz;
>  8: ADD temp[1].y, -const[1].zzzz, temp[1].xxxx;
>  9: MUL temp[0], temp[0], const[2].xxxx;
> 10: DP3 temp[1].x, temp[0], const[2].yyyy;
> 11: RCP temp[1].y, temp[1].yyyy;
> 12: ADD temp[1].x, temp[1], -const[1].zzzz;
> 13: MUL_SAT temp[1].x, temp[1], temp[1].yyyy;
> 14: MUL temp[1].y, -temp[1].xxxx, const[2].wwww;
> 15: ADD temp[1].y, temp[1], const[3].xxxx;
> 16: MUL temp[1].x, temp[1], temp[1];
> 17: MAD temp[1].x, temp[1], temp[1].yyyy, -const[2].zzzz;
> 18: MUL temp[1].x, temp[1], const[1].wwww;
> 19: ADD temp[1].x, temp[1], const[2].zzzz;
> 20: MOV temp[1].w, const[2].zzzz;
> 21: MUL temp[1].xyz, temp[1].xxxx, temp[0].wwww;
> 22: MUL output[0], temp[0], temp[1];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: TEX temp[0], input[1], RECT[0];
>  1: TEX temp[1], input[0], RECT[0];
>  2: ADD temp[2], temp[1], temp[0];
>  3: TEX temp[1], input[2], RECT[0];
>  4: ADD temp[1], temp[2], temp[1];
>  5: TEX temp[0], input[3], RECT[0];
>  6: ADD temp[0], temp[1], temp[0];
>  7: MOV temp[1].x, const[2].zzzz;
>  8: ADD temp[1].y, -const[1].zzzz, temp[1].xxxx;
>  9: MUL temp[0], temp[0], const[2].xxxx;
> 10: DP3 temp[1].x, temp[0], const[2].yyyy;
> 11: RCP temp[1].y, temp[1].yyyy;
> 12: ADD temp[1].x, temp[1], -const[1].zzzz;
> 13: MUL_SAT temp[1].x, temp[1], temp[1].yyyy;
> 14: MUL temp[1].y, -temp[1].xxxx, const[2].wwww;
> 15: ADD temp[1].y, temp[1], const[3].xxxx;
> 16: MUL temp[1].x, temp[1], temp[1];
> 17: MAD temp[1].x, temp[1], temp[1].yyyy, -const[2].zzzz;
> 18: MUL temp[1].x, temp[1], const[1].wwww;
> 19: ADD temp[1].x, temp[1], const[2].zzzz;
> 20: MOV temp[1].w, const[2].zzzz;
> 21: MUL temp[1].xyz, temp[1].xxxx, temp[0].wwww;
> 22: MUL output[0], temp[0], temp[1];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: TEX temp[0], input[1], RECT[0];
>  1: TEX temp[1], input[0], RECT[0];
>  2: ADD temp[2], temp[1], temp[0];
>  3: TEX temp[1], input[2], RECT[0];
>  4: ADD temp[1], temp[2], temp[1];
>  5: TEX temp[0], input[3], RECT[0];
>  6: ADD temp[0], temp[1], temp[0];
>  7: MOV temp[1].x, const[2].zzzz;
>  8: ADD temp[1].y, -const[1].zzzz, temp[1].xxxx;
>  9: MUL temp[0], temp[0], const[2].xxxx;
> 10: DP3 temp[1].x, temp[0], const[2].yyyy;
> 11: RCP temp[1].y, temp[1].yyyy;
> 12: ADD temp[1].x, temp[1], -const[1].zzzz;
> 13: MUL_SAT temp[1].x, temp[1], temp[1].yyyy;
> 14: MUL temp[1].y, -temp[1].xxxx, const[2].wwww;
> 15: ADD temp[1].y, temp[1], const[3].xxxx;
> 16: MUL temp[1].x, temp[1], temp[1];
> 17: MAD temp[1].x, temp[1], temp[1].yyyy, -const[2].zzzz;
> 18: MUL temp[1].x, temp[1], const[1].wwww;
> 19: ADD temp[1].x, temp[1], const[2].zzzz;
> 20: MOV temp[1].w, const[2].zzzz;
> 21: MUL temp[1].xyz, temp[1].xxxx, temp[0].wwww;
> 22: MUL output[0], temp[0], temp[1];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: TEX temp[0], input[1], RECT[0];
>  1: TEX temp[1], input[0], RECT[0];
>  2: ADD temp[2], temp[1], temp[0];
>  3: TEX temp[1], input[2], RECT[0];
>  4: ADD temp[1], temp[2], temp[1];
>  5: TEX temp[0], input[3], RECT[0];
>  6: ADD temp[0], temp[1], temp[0];
>  7: MOV temp[1].x, const[2].zzzz;
>  8: ADD temp[1].y, -const[1].zzzz, temp[1].xxxx;
>  9: MUL temp[0], temp[0], const[2].xxxx;
> 10: DP3 temp[1].x, temp[0], const[2].yyyy;
> 11: RCP temp[1].y, temp[1].yyyy;
> 12: ADD temp[1].x, temp[1], -const[1].zzzz;
> 13: MUL_SAT temp[1].x, temp[1], temp[1].yyyy;
> 14: MUL temp[1].y, -temp[1].xxxx, const[2].wwww;
> 15: ADD temp[1].y, temp[1], const[3].xxxx;
> 16: MUL temp[1].x, temp[1], temp[1];
> 17: MAD temp[1].x, temp[1], temp[1].yyyy, -const[2].zzzz;
> 18: MUL temp[1].x, temp[1], const[1].wwww;
> 19: ADD temp[1].x, temp[1], const[2].zzzz;
> 20: MOV temp[1].w, const[2].zzzz;
> 21: MUL temp[1].xyz, temp[1].xxxx, temp[0].wwww;
> 22: MUL output[0], temp[0], temp[1];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: MUL temp[3], input[1], const[4];
>  1: TEX temp[0], temp[3], 2D[0];
>  2: MUL temp[4], input[0], const[4];
>  3: TEX temp[1], temp[4], 2D[0];
>  4: ADD temp[2], temp[1], temp[0];
>  5: MUL temp[5], input[2], const[4];
>  6: TEX temp[1], temp[5], 2D[0];
>  7: ADD temp[1], temp[2], temp[1];
>  8: MUL temp[6], input[3], const[4];
>  9: TEX temp[0], temp[6], 2D[0];
> 10: ADD temp[0], temp[1], temp[0];
> 11: MOV temp[1].x, const[2].zzzz;
> 12: ADD temp[1].y, -const[1].zzzz, temp[1].xxxx;
> 13: MUL temp[0], temp[0], const[2].xxxx;
> 14: DP3 temp[1].x, temp[0], const[2].yyyy;
> 15: RCP temp[1].y, temp[1].yyyy;
> 16: ADD temp[1].x, temp[1], -const[1].zzzz;
> 17: MUL_SAT temp[1].x, temp[1], temp[1].yyyy;
> 18: MUL temp[1].y, -temp[1].xxxx, const[2].wwww;
> 19: ADD temp[1].y, temp[1], const[3].xxxx;
> 20: MUL temp[1].x, temp[1], temp[1];
> 21: MAD temp[1].x, temp[1], temp[1].yyyy, -const[2].zzzz;
> 22: MUL temp[1].x, temp[1], const[1].wwww;
> 23: ADD temp[1].x, temp[1], const[2].zzzz;
> 24: MOV temp[1].w, const[2].zzzz;
> 25: MUL temp[1].xyz, temp[1].xxxx, temp[0].wwww;
> 26: MUL output[0], temp[0], temp[1];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: MUL temp[3].xy, input[1].xy__, const[4].xy__;
>  1: TEX temp[0], temp[3].xy__, 2D[0];
>  2: MUL temp[4].xy, input[0].xy__, const[4].xy__;
>  3: TEX temp[1], temp[4].xy__, 2D[0];
>  4: ADD temp[2], temp[1], temp[0];
>  5: MUL temp[5].xy, input[2].xy__, const[4].xy__;
>  6: TEX temp[1], temp[5].xy__, 2D[0];
>  7: ADD temp[1], temp[2], temp[1];
>  8: MUL temp[6].xy, input[3].xy__, const[4].xy__;
>  9: TEX temp[0], temp[6].xy__, 2D[0];
> 10: ADD temp[0], temp[1], temp[0];
> 11: MOV temp[1].x, const[2].z___;
> 12: ADD temp[1].y, -const[1]._z__, temp[1]._x__;
> 13: MUL temp[0], temp[0], const[2].xxxx;
> 14: DP3 temp[1].x, temp[0].xyz_, const[2].yyy_;
> 15: RCP temp[1].y, temp[1].y___;
> 16: ADD temp[1].x, temp[1].x___, -const[1].z___;
> 17: MUL_SAT temp[1].x, temp[1].x___, temp[1].y___;
> 18: MUL temp[1].y, -temp[1]._x__, const[2]._w__;
> 19: ADD temp[1].y, temp[1]._y__, const[3]._x__;
> 20: MUL temp[1].x, temp[1].x___, temp[1].x___;
> 21: MAD temp[1].x, temp[1].x___, temp[1].y___, -const[2].z___;
> 22: MUL temp[1].x, temp[1].x___, const[1].w___;
> 23: ADD temp[1].x, temp[1].x___, const[2].z___;
> 24: MOV temp[1].w, const[2].___z;
> 25: MUL temp[1].xyz, temp[1].xxx_, temp[0].www_;
> 26: MUL output[0], temp[0], temp[1];
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: MUL temp[3].xy, input[1].xy__, const[4].xy__;
>  1: TEX temp[0], temp[3].xy__, 2D[0];
>  2: MUL temp[4].xy, input[0].xy__, const[4].xy__;
>  3: TEX temp[1], temp[4].xy__, 2D[0];
>  4: ADD temp[2], temp[1], temp[0];
>  5: MUL temp[5].xy, input[2].xy__, const[4].xy__;
>  6: TEX temp[1], temp[5].xy__, 2D[0];
>  7: ADD temp[1], temp[2], temp[1];
>  8: MUL temp[6].xy, input[3].xy__, const[4].xy__;
>  9: TEX temp[0], temp[6].xy__, 2D[0];
> 10: ADD temp[0], temp[1], temp[0];
> 11: ADD temp[1].y, -const[1]._z__, none._1__;
> 12: MUL temp[0], temp[0], const[2].xxxx;
> 13: DP3 temp[1].x, temp[0].xyz_, const[2].yyy_;
> 14: RCP temp[1].y, temp[1].y___;
> 15: ADD temp[1].x, temp[1].x___, -const[1].z___;
> 16: MUL_SAT temp[1].x, temp[1].x___, temp[1].y___;
> 17: MUL temp[1].y, -temp[1]._x__, const[2]._w__;
> 18: ADD temp[1].y, temp[1]._y__, const[3]._x__;
> 19: MUL temp[1].x, temp[1].x___, temp[1].x___;
> 20: MAD temp[1].x, temp[1].x___, temp[1].y___, -none.1___;
> 21: MUL temp[1].x, temp[1].x___, const[1].w___;
> 22: ADD temp[1].x, temp[1].x___, none.1___;
> 23: MOV temp[1].w, none.___1;
> 24: MUL temp[1].xyz, temp[1].xxx_, temp[0].www_;
> 25: MUL output[0], temp[0], temp[1];
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: MUL temp[3].xy, input[1].xy__, const[4].xy__;
>  1: TEX temp[0], temp[3].xy__, 2D[0];
>  2: MUL temp[4].xy, input[0].xy__, const[4].xy__;
>  3: TEX temp[1], temp[4].xy__, 2D[0];
>  4: ADD temp[2], temp[1], temp[0];
>  5: MUL temp[5].xy, input[2].xy__, const[4].xy__;
>  6: TEX temp[1], temp[5].xy__, 2D[0];
>  7: ADD temp[1], temp[2], temp[1];
>  8: MUL temp[6].xy, input[3].xy__, const[4].xy__;
>  9: TEX temp[0], temp[6].xy__, 2D[0];
> 10: ADD temp[0], temp[1], temp[0];
> 11: ADD temp[1].y, -const[1]._z__, none._1__;
> 12: MUL temp[0], temp[0], const[2].xxxx;
> 13: DP3 temp[1].x, temp[0].xyz_, const[2].yyy_;
> 14: RCP temp[1].y, temp[1].y___;
> 15: ADD temp[1].x, temp[1].x___, -const[1].z___;
> 16: MUL_SAT temp[1].x, temp[1].x___, temp[1].y___;
> 17: MUL temp[1].y, -temp[1]._x__, const[2]._w__;
> 18: ADD temp[1].y, temp[1]._y__, const[3]._x__;
> 19: MUL temp[1].x, temp[1].x___, temp[1].x___;
> 20: MAD temp[1].x, temp[1].x___, temp[1].y___, -none.1___;
> 21: MUL temp[1].x, temp[1].x___, const[1].w___;
> 22: ADD temp[1].x, temp[1].x___, none.1___;
> 23: MOV temp[1].w, none.___1;
> 24: MUL temp[1].xyz, temp[1].xxx_, temp[0].www_;
> 25: MUL output[0], temp[0], temp[1];
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: src0.xyz = input[1], src1.xyz = const[4]
>     MAD temp[3].xy, src0.xy_, src1.xy_, src0.000
>  1: TEX temp[0], temp[3].xy__, 2D[0];
>  2: src0.xyz = input[0], src1.xyz = const[4]
>     MAD temp[4].xy, src0.xy_, src1.xy_, src0.000
>  3: TEX temp[1], temp[4].xy__, 2D[0];
>  4: src0.xyz = temp[1], src0.w = temp[1], src1.xyz = temp[0], src1.w = temp[0]
>     MAD temp[2].xyz, src0.xyz, src0.111, src1.xyz
>     MAD temp[2].w, src0.w, src0.1, src1.w
>  5: src0.xyz = input[2], src1.xyz = const[4]
>     MAD temp[5].xy, src0.xy_, src1.xy_, src0.000
>  6: TEX temp[1], temp[5].xy__, 2D[0];
>  7: src0.xyz = temp[2], src0.w = temp[2], src1.xyz = temp[1], src1.w = temp[1]
>     MAD temp[1].xyz, src0.xyz, src0.111, src1.xyz
>     MAD temp[1].w, src0.w, src0.1, src1.w
>  8: src0.xyz = input[3], src1.xyz = const[4]
>     MAD temp[6].xy, src0.xy_, src1.xy_, src0.000
>  9: TEX temp[0], temp[6].xy__, 2D[0];
> 10: src0.xyz = temp[1], src0.w = temp[1], src1.xyz = temp[0], src1.w = temp[0]
>     MAD temp[0].xyz, src0.xyz, src0.111, src1.xyz
>     MAD temp[0].w, src0.w, src0.1, src1.w
> 11: src0.xyz = const[1]
>     MAD temp[1].y, -src0._z_, src0.111, src0._1_
> 12: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[2]
>     MAD temp[0].xyz, src0.xyz, src1.xxx, src0.000
>     MAD temp[0].w, src0.w, src1.x, src0.0
> 13: src0.xyz = temp[0], src1.xyz = const[2]
>     DP3 temp[1].x, src0.xyz, src1.yyy
> 14: src0.xyz = temp[1]
>     REPL_ALPHA temp[1].y
>     RCP, src0.y
> 15: src0.xyz = temp[1], src1.xyz = const[1]
>     MAD temp[1].x, src0.x__, src0.111, -src1.z__
> 16: src0.xyz = temp[1]
>     MAD_SAT temp[1].x, src0.x__, src0.y__, src0.000
> 17: src0.xyz = temp[1], src0.w = const[2]
>     MAD temp[1].y, -src0._x_, src0._w_, src0.000
> 18: src0.xyz = temp[1], src1.xyz = const[3]
>     MAD temp[1].y, src0._y_, src0.111, src1._x_
> 19: src0.xyz = temp[1]
>     MAD temp[1].x, src0.x__, src0.x__, src0.000
> 20: src0.xyz = temp[1]
>     MAD temp[1].x, src0.x__, src0.y__, -src0.1__
> 21: src0.xyz = temp[1], src0.w = const[1]
>     MAD temp[1].x, src0.x__, src0.w__, src0.000
> 22: src0.xyz = temp[1]
>     MAD temp[1].x, src0.x__, src0.111, src0.1__
> 23: 
>     MAD temp[1].w, src0.1, src0.1, src0.0
> 24: src0.xyz = temp[1], src0.w = temp[0]
>     MAD temp[1].xyz, src0.xxx, src0.www, src0.000
> 25: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = temp[1], src1.w = temp[1]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: src0.xyz = input[3], src1.xyz = const[4]
>     MAD temp[6].xy, src0.xy_, src1.xy_, src0.000
>  1: src0.xyz = input[2], src1.xyz = const[4]
>     MAD temp[5].xy, src0.xy_, src1.xy_, src0.000
>  2: src0.xyz = input[0], src1.xyz = const[4]
>     MAD temp[4].xy, src0.xy_, src1.xy_, src0.000
>  3: src0.xyz = input[1], src1.xyz = const[4]
>     MAD temp[3].xy, src0.xy_, src1.xy_, src0.000
>  4: BEGIN_TEX;
>  5: TEX temp[0], temp[3].xy__, 2D[0];
>  6: TEX temp[1], temp[4].xy__, 2D[0];
>  7: src0.xyz = temp[1], src0.w = temp[1], src1.xyz = temp[0], src1.w = temp[0]
>     MAD temp[2].xyz, src0.xyz, src0.111, src1.xyz
>     MAD temp[2].w, src0.w, src0.1, src1.w
>  8: BEGIN_TEX;
>  9: TEX temp[0], temp[6].xy__, 2D[0];
> 10: TEX temp[1], temp[5].xy__, 2D[0];
> 11: src0.xyz = temp[2], src0.w = temp[2], src1.xyz = temp[1], src1.w = temp[1]
>     MAD temp[1].xyz, src0.xyz, src0.111, src1.xyz
>     MAD temp[1].w, src0.w, src0.1, src1.w
> 12: src0.xyz = temp[1], src0.w = temp[1], src1.xyz = temp[0], src1.w = temp[0]
>     MAD temp[0].xyz, src0.xyz, src0.111, src1.xyz
>     MAD temp[0].w, src0.w, src0.1, src1.w
> 13: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[2]
>     MAD temp[0].xyz, src0.xyz, src1.xxx, src0.000
>     MAD temp[0].w, src0.w, src1.x, src0.0
> 14: src0.xyz = temp[0], src1.xyz = const[2]
>     DP3 temp[1].x, src0.xyz, src1.yyy
>     MAD temp[1].w, src0.1, src0.1, src0.0
> 15: src0.xyz = temp[1], src1.xyz = const[1]
>     MAD temp[1].x, src0.x__, src0.111, -src1.z__
> 16: src0.xyz = const[1]
>     MAD temp[1].y, -src0._z_, src0.111, src0._1_
> 17: src0.xyz = temp[1]
>     REPL_ALPHA temp[1].y
>     RCP, src0.y
> 18: src0.xyz = temp[1]
>     MAD_SAT temp[1].x, src0.x__, src0.y__, src0.000
> 19: src0.xyz = temp[1], src0.w = const[2]
>     MAD temp[1].y, -src0._x_, src0._w_, src0.000
> 20: src0.xyz = temp[1], src1.xyz = const[3]
>     MAD temp[1].y, src0._y_, src0.111, src1._x_
> 21: src0.xyz = temp[1]
>     MAD temp[1].x, src0.x__, src0.x__, src0.000
> 22: src0.xyz = temp[1]
>     MAD temp[1].x, src0.x__, src0.y__, -src0.1__
> 23: src0.xyz = temp[1], src0.w = const[1]
>     MAD temp[1].x, src0.x__, src0.w__, src0.000
> 24: src0.xyz = temp[1]
>     MAD temp[1].x, src0.x__, src0.111, src0.1__
> 25: src0.xyz = temp[1], src0.w = temp[0]
>     MAD temp[1].xyz, src0.xxx, src0.www, src0.000
> 26: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = temp[1], src1.w = temp[1]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: src0.xyz = temp[3], src1.xyz = const[4]
>     MAD temp[4].xy, src0.xy_, src1.xy_, src0.000
>  1: src0.xyz = temp[2], src1.xyz = const[4]
>     MAD temp[3].xy, src0.xy_, src1.xy_, src0.000
>  2: src0.xyz = temp[0], src1.xyz = const[4]
>     MAD temp[2].xy, src0.xy_, src1.xy_, src0.000
>  3: src0.xyz = temp[1], src1.xyz = const[4]
>     MAD temp[0].xy, src0.xy_, src1.xy_, src0.000
>  4: BEGIN_TEX;
>  5: TEX temp[0], temp[0].xy__, 2D[0];
>  6: TEX temp[1], temp[2].xy__, 2D[0];
>  7: src0.xyz = temp[1], src0.w = temp[1], src1.xyz = temp[0], src1.w = temp[0]
>     MAD temp[2].xyz, src0.xyz, src0.111, src1.xyz
>     MAD temp[2].w, src0.w, src0.1, src1.w
>  8: BEGIN_TEX;
>  9: TEX temp[0], temp[4].xy__, 2D[0];
> 10: TEX temp[1], temp[3].xy__, 2D[0];
> 11: src0.xyz = temp[2], src0.w = temp[2], src1.xyz = temp[1], src1.w = temp[1]
>     MAD temp[1].xyz, src0.xyz, src0.111, src1.xyz
>     MAD temp[1].w, src0.w, src0.1, src1.w
> 12: src0.xyz = temp[1], src0.w = temp[1], src1.xyz = temp[0], src1.w = temp[0]
>     MAD temp[0].xyz, src0.xyz, src0.111, src1.xyz
>     MAD temp[0].w, src0.w, src0.1, src1.w
> 13: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[2]
>     MAD temp[0].xyz, src0.xyz, src1.xxx, src0.000
>     MAD temp[0].w, src0.w, src1.x, src0.0
> 14: src0.xyz = temp[0], src1.xyz = const[2]
>     DP3 temp[1].x, src0.xyz, src1.yyy
>     MAD temp[1].w, src0.1, src0.1, src0.0
> 15: src0.xyz = temp[1], src1.xyz = const[1]
>     MAD temp[1].x, src0.x__, src0.111, -src1.z__
> 16: src0.xyz = const[1]
>     MAD temp[1].y, -src0._z_, src0.111, src0._1_
> 17: src0.xyz = temp[1]
>     REPL_ALPHA temp[1].y
>     RCP, src0.y
> 18: src0.xyz = temp[1]
>     MAD_SAT temp[1].x, src0.x__, src0.y__, src0.000
> 19: src0.xyz = temp[1], src0.w = const[2]
>     MAD temp[1].y, -src0._x_, src0._w_, src0.000
> 20: src0.xyz = temp[1], src1.xyz = const[3]
>     MAD temp[1].y, src0._y_, src0.111, src1._x_
> 21: src0.xyz = temp[1]
>     MAD temp[1].x, src0.x__, src0.x__, src0.000
> 22: src0.xyz = temp[1]
>     MAD temp[1].x, src0.x__, src0.y__, -src0.1__
> 23: src0.xyz = temp[1], src0.w = const[1]
>     MAD temp[1].x, src0.x__, src0.w__, src0.000
> 24: src0.xyz = temp[1]
>     MAD temp[1].x, src0.x__, src0.111, src0.1__
> 25: src0.xyz = temp[1], src0.w = temp[0]
>     MAD temp[1].xyz, src0.xxx, src0.www, src0.000
> 26: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = temp[1], src1.w = temp[1]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>pc=15*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 3, tex_end: 0  (code_addr: 000000c0)
>  0: xyz:  t3  c4  t0 -> t4.xy                (01900903)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t3.xyz   c4.xyz      0.0    op: 00050200
>       w:     t3.x     t3.x     t3.x    op: 00000000
>  1: xyz:  t2  c4  t0 -> t3.xy                (018c0902)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz   c4.xyz      0.0    op: 00050200
>       w:     t2.x     t2.x     t2.x    op: 00000000
>  2: xyz:  t0  c4  t0 -> t2.xy                (01880900)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xyz   c4.xyz      0.0    op: 00050200
>       w:     t0.x     t0.x     t0.x    op: 00000000
>  3: xyz:  t1  c4  t0 -> t0.xy                (01800901)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   c4.xyz      0.0    op: 00050200
>       w:     t1.x     t1.x     t1.x    op: 00000000
>NODE 1: alu_offset: 4, tex_offset: 0, alu_end: 0, tex_end: 1  (code_addr: 00020004)
>  TEX:
>    TEX t0, t0, texture[0]   (00008000)
>    TEX t1, t2, texture[0]   (00008042)
>  4: xyz:  t1  t0  t0 -> t2.xyz               (03880001)
>       w:  t1  t0  t0 -> t2.w                 (00880001)
>     xyz:   t1.xyz      1.0   t0.xyz    op: 00010a80
>       w:     t1.w      1.0     t0.w    op: 00028889
>NODE 2: alu_offset: 5, tex_offset: 2, alu_end: 15, tex_end: 1  (code_addr: 004223c5)
>  TEX:
>    TEX t0, t4, texture[0]   (00008004)
>    TEX t1, t3, texture[0]   (00008043)
>  5: xyz:  t2  t1  t0 -> t1.xyz               (03840042)
>       w:  t2  t1  t0 -> t1.w                 (00840042)
>     xyz:   t2.xyz      1.0   t1.xyz    op: 00010a80
>       w:     t2.w      1.0     t1.w    op: 00028889
>  6: xyz:  t1  t0  t0 -> t0.xyz               (03800001)
>       w:  t1  t0  t0 -> t0.w                 (00800001)
>     xyz:   t1.xyz      1.0   t0.xyz    op: 00010a80
>       w:     t1.w      1.0     t0.w    op: 00028889
>  7: xyz:  t0  c2  t0 -> t0.xyz               (03800880)
>       w:  t0  t0  t0 -> t0.w                 (00800000)
>     xyz:   t0.xyz   c2.xxx      0.0    op: 00050280
>       w:     t0.w     c2.x      0.0    op: 00040189
>  8: xyz:  t0  c2  t0 -> t1.x                 (00840880)
>       w:  t0  t0  t0 -> t1.w                 (00840000)
>     xyz:   t0.xyz   c2.yyy   t0.xxx    op: 00804300
>       w:      1.0      1.0      0.0    op: 00040891
>  9: xyz:  t1  c1  t0 -> t1.x                 (00840841)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz      1.0  -c1.zzz    op: 0009ca80
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 10: xyz:  c1  t0  t0 -> t1.y                 (01040021)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:  -c1.zzz      1.0      1.0    op: 00054aa3
>       w:     c1.x     c1.x     c1.x    op: 00000000
> 11: xyz:  t1  t0  t0 -> t1.y                 (01040001)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xxx   t1.xxx   t1.xxx    op: 05004081
>       w:     t1.y     t1.x     t1.x    op: 05000001
> 12: xyz:  t1  t0  t0 -> t1.x                 (00840001)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   t1.yyy      0.0    op: 40050100
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 13: xyz:  t1  t0  t0 -> t1.y                 (01040001)
>       w:  c2  t0  t0 ->                      (00000022)
>     xyz:  -t1.xxx   c2.www      0.0    op: 00050621
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 14: xyz:  t1  c3  t0 -> t1.y                 (010408c1)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz      1.0   c3.xxx    op: 00014a80
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 15: xyz:  t1  t0  t0 -> t1.x                 (00840001)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   t1.xyz      0.0    op: 00050000
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 16: xyz:  t1  t0  t0 -> t1.x                 (00840001)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   t1.yyy     -1.0    op: 000d4100
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 17: xyz:  t1  t0  t0 -> t1.x                 (00840001)
>       w:  c1  t0  t0 ->                      (00000021)
>     xyz:   t1.xyz   c1.www      0.0    op: 00050600
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 18: xyz:  t1  t0  t0 -> t1.x                 (00840001)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz      1.0      1.0    op: 00054a80
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 19: xyz:  t1  t0  t0 -> t1.xyz               (03840001)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xxx   t0.www      0.0    op: 00050601
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 20: xyz:  t0  t1  t0 -> o0.xyz               (1c000040)
>       w:  t0  t1  t0 -> o0.w                 (01000040)
>     xyz:   t0.xyz   t1.xyz      0.0    op: 00050200
>       w:     t0.w     t1.w      0.0    op: 00040509
>r300: Initial fragment program
>FRAG
>DCL IN[0], GENERIC[0], PERSPECTIVE
>DCL IN[1], GENERIC[1], PERSPECTIVE
>DCL IN[2], GENERIC[2], PERSPECTIVE
>DCL IN[3], GENERIC[3], PERSPECTIVE
>DCL IN[4], GENERIC[4], PERSPECTIVE
>DCL IN[5], GENERIC[5], PERSPECTIVE
>DCL IN[6], GENERIC[6], PERSPECTIVE
>DCL IN[7], GENERIC[7], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL TEMP[0..1]
>IMM FLT32 {    0.0714,     0.1071,     0.1429,     0.1786}
>  0: TEX TEMP[0], IN[1], SAMP[0], RECT
>  1: MUL TEMP[1], TEMP[0], IMM[0].yyyy
>  2: TEX TEMP[0], IN[0], SAMP[0], RECT
>  3: MAD TEMP[1], TEMP[0], IMM[0].xxxx, TEMP[1]
>  4: TEX TEMP[0], IN[2], SAMP[0], RECT
>  5: MAD TEMP[1], TEMP[0], IMM[0].zzzz, TEMP[1]
>  6: TEX TEMP[0], IN[3], SAMP[0], RECT
>  7: MAD TEMP[1], TEMP[0], IMM[0].wwww, TEMP[1]
>  8: TEX TEMP[0], IN[4], SAMP[0], RECT
>  9: MAD TEMP[1], TEMP[0], IMM[0].wwww, TEMP[1]
> 10: TEX TEMP[0], IN[5], SAMP[0], RECT
> 11: MAD TEMP[1], TEMP[0], IMM[0].zzzz, TEMP[1]
> 12: TEX TEMP[0], IN[6], SAMP[0], RECT
> 13: MAD TEMP[1], TEMP[0], IMM[0].yyyy, TEMP[1]
> 14: TEX TEMP[0], IN[7], SAMP[0], RECT
> 15: MAD OUT[0], TEMP[0], IMM[0].xxxx, TEMP[1]
> 16: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: TEX temp[0], input[1], RECT[0];
>  1: MUL temp[1], temp[0], const[0].yyyy;
>  2: TEX temp[0], input[0], RECT[0];
>  3: MAD temp[1], temp[0], const[0].xxxx, temp[1];
>  4: TEX temp[0], input[2], RECT[0];
>  5: MAD temp[1], temp[0], const[0].zzzz, temp[1];
>  6: TEX temp[0], input[3], RECT[0];
>  7: MAD temp[1], temp[0], const[0].wwww, temp[1];
>  8: TEX temp[0], input[4], RECT[0];
>  9: MAD temp[1], temp[0], const[0].wwww, temp[1];
> 10: TEX temp[0], input[5], RECT[0];
> 11: MAD temp[1], temp[0], const[0].zzzz, temp[1];
> 12: TEX temp[0], input[6], RECT[0];
> 13: MAD temp[1], temp[0], const[0].yyyy, temp[1];
> 14: TEX temp[0], input[7], RECT[0];
> 15: MAD output[0], temp[0], const[0].xxxx, temp[1];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: TEX temp[0], input[1], RECT[0];
>  1: MUL temp[1], temp[0], const[0].yyyy;
>  2: TEX temp[0], input[0], RECT[0];
>  3: MAD temp[1], temp[0], const[0].xxxx, temp[1];
>  4: TEX temp[0], input[2], RECT[0];
>  5: MAD temp[1], temp[0], const[0].zzzz, temp[1];
>  6: TEX temp[0], input[3], RECT[0];
>  7: MAD temp[1], temp[0], const[0].wwww, temp[1];
>  8: TEX temp[0], input[4], RECT[0];
>  9: MAD temp[1], temp[0], const[0].wwww, temp[1];
> 10: TEX temp[0], input[5], RECT[0];
> 11: MAD temp[1], temp[0], const[0].zzzz, temp[1];
> 12: TEX temp[0], input[6], RECT[0];
> 13: MAD temp[1], temp[0], const[0].yyyy, temp[1];
> 14: TEX temp[0], input[7], RECT[0];
> 15: MAD output[0], temp[0], const[0].xxxx, temp[1];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: TEX temp[0], input[1], RECT[0];
>  1: MUL temp[1], temp[0], const[0].yyyy;
>  2: TEX temp[0], input[0], RECT[0];
>  3: MAD temp[1], temp[0], const[0].xxxx, temp[1];
>  4: TEX temp[0], input[2], RECT[0];
>  5: MAD temp[1], temp[0], const[0].zzzz, temp[1];
>  6: TEX temp[0], input[3], RECT[0];
>  7: MAD temp[1], temp[0], const[0].wwww, temp[1];
>  8: TEX temp[0], input[4], RECT[0];
>  9: MAD temp[1], temp[0], const[0].wwww, temp[1];
> 10: TEX temp[0], input[5], RECT[0];
> 11: MAD temp[1], temp[0], const[0].zzzz, temp[1];
> 12: TEX temp[0], input[6], RECT[0];
> 13: MAD temp[1], temp[0], const[0].yyyy, temp[1];
> 14: TEX temp[0], input[7], RECT[0];
> 15: MAD output[0], temp[0], const[0].xxxx, temp[1];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: TEX temp[0], input[1], RECT[0];
>  1: MUL temp[1], temp[0], const[0].yyyy;
>  2: TEX temp[0], input[0], RECT[0];
>  3: MAD temp[1], temp[0], const[0].xxxx, temp[1];
>  4: TEX temp[0], input[2], RECT[0];
>  5: MAD temp[1], temp[0], const[0].zzzz, temp[1];
>  6: TEX temp[0], input[3], RECT[0];
>  7: MAD temp[1], temp[0], const[0].wwww, temp[1];
>  8: TEX temp[0], input[4], RECT[0];
>  9: MAD temp[1], temp[0], const[0].wwww, temp[1];
> 10: TEX temp[0], input[5], RECT[0];
> 11: MAD temp[1], temp[0], const[0].zzzz, temp[1];
> 12: TEX temp[0], input[6], RECT[0];
> 13: MAD temp[1], temp[0], const[0].yyyy, temp[1];
> 14: TEX temp[0], input[7], RECT[0];
> 15: MAD output[0], temp[0], const[0].xxxx, temp[1];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: MUL temp[2], input[1], const[1];
>  1: TEX temp[0], temp[2], 2D[0];
>  2: MUL temp[1], temp[0], const[0].yyyy;
>  3: MUL temp[3], input[0], const[1];
>  4: TEX temp[0], temp[3], 2D[0];
>  5: MAD temp[1], temp[0], const[0].xxxx, temp[1];
>  6: MUL temp[4], input[2], const[1];
>  7: TEX temp[0], temp[4], 2D[0];
>  8: MAD temp[1], temp[0], const[0].zzzz, temp[1];
>  9: MUL temp[5], input[3], const[1];
> 10: TEX temp[0], temp[5], 2D[0];
> 11: MAD temp[1], temp[0], const[0].wwww, temp[1];
> 12: MUL temp[6], input[4], const[1];
> 13: TEX temp[0], temp[6], 2D[0];
> 14: MAD temp[1], temp[0], const[0].wwww, temp[1];
> 15: MUL temp[7], input[5], const[1];
> 16: TEX temp[0], temp[7], 2D[0];
> 17: MAD temp[1], temp[0], const[0].zzzz, temp[1];
> 18: MUL temp[8], input[6], const[1];
> 19: TEX temp[0], temp[8], 2D[0];
> 20: MAD temp[1], temp[0], const[0].yyyy, temp[1];
> 21: MUL temp[9], input[7], const[1];
> 22: TEX temp[0], temp[9], 2D[0];
> 23: MAD output[0], temp[0], const[0].xxxx, temp[1];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: MUL temp[2].xy, input[1].xy__, const[1].xy__;
>  1: TEX temp[0], temp[2].xy__, 2D[0];
>  2: MUL temp[1], temp[0], const[0].yyyy;
>  3: MUL temp[3].xy, input[0].xy__, const[1].xy__;
>  4: TEX temp[0], temp[3].xy__, 2D[0];
>  5: MAD temp[1], temp[0], const[0].xxxx, temp[1];
>  6: MUL temp[4].xy, input[2].xy__, const[1].xy__;
>  7: TEX temp[0], temp[4].xy__, 2D[0];
>  8: MAD temp[1], temp[0], const[0].zzzz, temp[1];
>  9: MUL temp[5].xy, input[3].xy__, const[1].xy__;
> 10: TEX temp[0], temp[5].xy__, 2D[0];
> 11: MAD temp[1], temp[0], const[0].wwww, temp[1];
> 12: MUL temp[6].xy, input[4].xy__, const[1].xy__;
> 13: TEX temp[0], temp[6].xy__, 2D[0];
> 14: MAD temp[1], temp[0], const[0].wwww, temp[1];
> 15: MUL temp[7].xy, input[5].xy__, const[1].xy__;
> 16: TEX temp[0], temp[7].xy__, 2D[0];
> 17: MAD temp[1], temp[0], const[0].zzzz, temp[1];
> 18: MUL temp[8].xy, input[6].xy__, const[1].xy__;
> 19: TEX temp[0], temp[8].xy__, 2D[0];
> 20: MAD temp[1], temp[0], const[0].yyyy, temp[1];
> 21: MUL temp[9].xy, input[7].xy__, const[1].xy__;
> 22: TEX temp[0], temp[9].xy__, 2D[0];
> 23: MAD output[0], temp[0], const[0].xxxx, temp[1];
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: MUL temp[2].xy, input[1].xy__, const[1].xy__;
>  1: TEX temp[0], temp[2].xy__, 2D[0];
>  2: MUL temp[1], temp[0], const[0].yyyy;
>  3: MUL temp[3].xy, input[0].xy__, const[1].xy__;
>  4: TEX temp[0], temp[3].xy__, 2D[0];
>  5: MAD temp[1], temp[0], const[0].xxxx, temp[1];
>  6: MUL temp[4].xy, input[2].xy__, const[1].xy__;
>  7: TEX temp[0], temp[4].xy__, 2D[0];
>  8: MAD temp[1], temp[0], const[0].zzzz, temp[1];
>  9: MUL temp[5].xy, input[3].xy__, const[1].xy__;
> 10: TEX temp[0], temp[5].xy__, 2D[0];
> 11: MAD temp[1], temp[0], const[0].wwww, temp[1];
> 12: MUL temp[6].xy, input[4].xy__, const[1].xy__;
> 13: TEX temp[0], temp[6].xy__, 2D[0];
> 14: MAD temp[1], temp[0], const[0].wwww, temp[1];
> 15: MUL temp[7].xy, input[5].xy__, const[1].xy__;
> 16: TEX temp[0], temp[7].xy__, 2D[0];
> 17: MAD temp[1], temp[0], const[0].zzzz, temp[1];
> 18: MUL temp[8].xy, input[6].xy__, const[1].xy__;
> 19: TEX temp[0], temp[8].xy__, 2D[0];
> 20: MAD temp[1], temp[0], const[0].yyyy, temp[1];
> 21: MUL temp[9].xy, input[7].xy__, const[1].xy__;
> 22: TEX temp[0], temp[9].xy__, 2D[0];
> 23: MAD output[0], temp[0], const[0].xxxx, temp[1];
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: MUL temp[2].xy, input[1].xy__, const[1].xy__;
>  1: TEX temp[0], temp[2].xy__, 2D[0];
>  2: MUL temp[1], temp[0], const[0].yyyy;
>  3: MUL temp[3].xy, input[0].xy__, const[1].xy__;
>  4: TEX temp[0], temp[3].xy__, 2D[0];
>  5: MAD temp[1], temp[0], const[0].xxxx, temp[1];
>  6: MUL temp[4].xy, input[2].xy__, const[1].xy__;
>  7: TEX temp[0], temp[4].xy__, 2D[0];
>  8: MAD temp[1], temp[0], const[0].zzzz, temp[1];
>  9: MUL temp[5].xy, input[3].xy__, const[1].xy__;
> 10: TEX temp[0], temp[5].xy__, 2D[0];
> 11: MAD temp[1], temp[0], const[0].wwww, temp[1];
> 12: MUL temp[6].xy, input[4].xy__, const[1].xy__;
> 13: TEX temp[0], temp[6].xy__, 2D[0];
> 14: MAD temp[1], temp[0], const[0].wwww, temp[1];
> 15: MUL temp[7].xy, input[5].xy__, const[1].xy__;
> 16: TEX temp[0], temp[7].xy__, 2D[0];
> 17: MAD temp[1], temp[0], const[0].zzzz, temp[1];
> 18: MUL temp[8].xy, input[6].xy__, const[1].xy__;
> 19: TEX temp[0], temp[8].xy__, 2D[0];
> 20: MAD temp[1], temp[0], const[0].yyyy, temp[1];
> 21: MUL temp[9].xy, input[7].xy__, const[1].xy__;
> 22: TEX temp[0], temp[9].xy__, 2D[0];
> 23: MAD output[0], temp[0], const[0].xxxx, temp[1];
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: src0.xyz = input[1], src1.xyz = const[1]
>     MAD temp[2].xy, src0.xy_, src1.xy_, src0.000
>  1: TEX temp[0], temp[2].xy__, 2D[0];
>  2: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[0]
>     MAD temp[1].xyz, src0.xyz, src1.yyy, src0.000
>     MAD temp[1].w, src0.w, src1.y, src0.0
>  3: src0.xyz = input[0], src1.xyz = const[1]
>     MAD temp[3].xy, src0.xy_, src1.xy_, src0.000
>  4: TEX temp[0], temp[3].xy__, 2D[0];
>  5: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[0], src1.w = temp[1], src2.xyz = temp[1]
>     MAD temp[1].xyz, src0.xyz, src1.xxx, src2.xyz
>     MAD temp[1].w, src0.w, src1.x, src1.w
>  6: src0.xyz = input[2], src1.xyz = const[1]
>     MAD temp[4].xy, src0.xy_, src1.xy_, src0.000
>  7: TEX temp[0], temp[4].xy__, 2D[0];
>  8: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[0], src1.w = temp[1], src2.xyz = temp[1]
>     MAD temp[1].xyz, src0.xyz, src1.zzz, src2.xyz
>     MAD temp[1].w, src0.w, src1.z, src1.w
>  9: src0.xyz = input[3], src1.xyz = const[1]
>     MAD temp[5].xy, src0.xy_, src1.xy_, src0.000
> 10: TEX temp[0], temp[5].xy__, 2D[0];
> 11: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = temp[1], src1.w = const[0], src2.w = temp[1]
>     MAD temp[1].xyz, src0.xyz, src1.www, src1.xyz
>     MAD temp[1].w, src0.w, src1.w, src2.w
> 12: src0.xyz = input[4], src1.xyz = const[1]
>     MAD temp[6].xy, src0.xy_, src1.xy_, src0.000
> 13: TEX temp[0], temp[6].xy__, 2D[0];
> 14: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = temp[1], src1.w = const[0], src2.w = temp[1]
>     MAD temp[1].xyz, src0.xyz, src1.www, src1.xyz
>     MAD temp[1].w, src0.w, src1.w, src2.w
> 15: src0.xyz = input[5], src1.xyz = const[1]
>     MAD temp[7].xy, src0.xy_, src1.xy_, src0.000
> 16: TEX temp[0], temp[7].xy__, 2D[0];
> 17: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[0], src1.w = temp[1], src2.xyz = temp[1]
>     MAD temp[1].xyz, src0.xyz, src1.zzz, src2.xyz
>     MAD temp[1].w, src0.w, src1.z, src1.w
> 18: src0.xyz = input[6], src1.xyz = const[1]
>     MAD temp[8].xy, src0.xy_, src1.xy_, src0.000
> 19: TEX temp[0], temp[8].xy__, 2D[0];
> 20: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[0], src1.w = temp[1], src2.xyz = temp[1]
>     MAD temp[1].xyz, src0.xyz, src1.yyy, src2.xyz
>     MAD temp[1].w, src0.w, src1.y, src1.w
> 21: src0.xyz = input[7], src1.xyz = const[1]
>     MAD temp[9].xy, src0.xy_, src1.xy_, src0.000
> 22: TEX temp[0], temp[9].xy__, 2D[0];
> 23: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[0], src1.w = temp[1], src2.xyz = temp[1]
>     MAD color[0].xyz, src0.xyz, src1.xxx, src2.xyz
>     MAD color[0].w, src0.w, src1.x, src1.w
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: src0.xyz = input[7], src1.xyz = const[1]
>     MAD temp[9].xy, src0.xy_, src1.xy_, src0.000
>  1: src0.xyz = input[6], src1.xyz = const[1]
>     MAD temp[8].xy, src0.xy_, src1.xy_, src0.000
>  2: src0.xyz = input[5], src1.xyz = const[1]
>     MAD temp[7].xy, src0.xy_, src1.xy_, src0.000
>  3: src0.xyz = input[4], src1.xyz = const[1]
>     MAD temp[6].xy, src0.xy_, src1.xy_, src0.000
>  4: src0.xyz = input[3], src1.xyz = const[1]
>     MAD temp[5].xy, src0.xy_, src1.xy_, src0.000
>  5: src0.xyz = input[2], src1.xyz = const[1]
>     MAD temp[4].xy, src0.xy_, src1.xy_, src0.000
>  6: src0.xyz = input[0], src1.xyz = const[1]
>     MAD temp[3].xy, src0.xy_, src1.xy_, src0.000
>  7: src0.xyz = input[1], src1.xyz = const[1]
>     MAD temp[2].xy, src0.xy_, src1.xy_, src0.000
>  8: BEGIN_TEX;
>  9: TEX temp[0], temp[2].xy__, 2D[0];
> 10: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[0]
>     MAD temp[1].xyz, src0.xyz, src1.yyy, src0.000
>     MAD temp[1].w, src0.w, src1.y, src0.0
> 11: BEGIN_TEX;
> 12: TEX temp[0], temp[3].xy__, 2D[0];
> 13: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[0], src1.w = temp[1], src2.xyz = temp[1]
>     MAD temp[1].xyz, src0.xyz, src1.xxx, src2.xyz
>     MAD temp[1].w, src0.w, src1.x, src1.w
> 14: BEGIN_TEX;
> 15: TEX temp[0], temp[4].xy__, 2D[0];
> 16: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[0], src1.w = temp[1], src2.xyz = temp[1]
>     MAD temp[1].xyz, src0.xyz, src1.zzz, src2.xyz
>     MAD temp[1].w, src0.w, src1.z, src1.w
> 17: BEGIN_TEX;
> 18: TEX temp[0], temp[5].xy__, 2D[0];
> 19: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = temp[1], src1.w = const[0], src2.w = temp[1]
>     MAD temp[1].xyz, src0.xyz, src1.www, src1.xyz
>     MAD temp[1].w, src0.w, src1.w, src2.w
> 20: BEGIN_TEX;
> 21: TEX temp[0], temp[6].xy__, 2D[0];
> 22: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = temp[1], src1.w = const[0], src2.w = temp[1]
>     MAD temp[1].xyz, src0.xyz, src1.www, src1.xyz
>     MAD temp[1].w, src0.w, src1.w, src2.w
> 23: BEGIN_TEX;
> 24: TEX temp[0], temp[7].xy__, 2D[0];
> 25: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[0], src1.w = temp[1], src2.xyz = temp[1]
>     MAD temp[1].xyz, src0.xyz, src1.zzz, src2.xyz
>     MAD temp[1].w, src0.w, src1.z, src1.w
> 26: BEGIN_TEX;
> 27: TEX temp[0], temp[8].xy__, 2D[0];
> 28: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[0], src1.w = temp[1], src2.xyz = temp[1]
>     MAD temp[1].xyz, src0.xyz, src1.yyy, src2.xyz
>     MAD temp[1].w, src0.w, src1.y, src1.w
> 29: BEGIN_TEX;
> 30: TEX temp[0], temp[9].xy__, 2D[0];
> 31: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[0], src1.w = temp[1], src2.xyz = temp[1]
>     MAD color[0].xyz, src0.xyz, src1.xxx, src2.xyz
>     MAD color[0].w, src0.w, src1.x, src1.w
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: src0.xyz = temp[7], src1.xyz = const[1]
>     MAD temp[8].xy, src0.xy_, src1.xy_, src0.000
>  1: src0.xyz = temp[6], src1.xyz = const[1]
>     MAD temp[7].xy, src0.xy_, src1.xy_, src0.000
>  2: src0.xyz = temp[5], src1.xyz = const[1]
>     MAD temp[6].xy, src0.xy_, src1.xy_, src0.000
>  3: src0.xyz = temp[4], src1.xyz = const[1]
>     MAD temp[5].xy, src0.xy_, src1.xy_, src0.000
>  4: src0.xyz = temp[3], src1.xyz = const[1]
>     MAD temp[4].xy, src0.xy_, src1.xy_, src0.000
>  5: src0.xyz = temp[2], src1.xyz = const[1]
>     MAD temp[3].xy, src0.xy_, src1.xy_, src0.000
>  6: src0.xyz = temp[0], src1.xyz = const[1]
>     MAD temp[2].xy, src0.xy_, src1.xy_, src0.000
>  7: src0.xyz = temp[1], src1.xyz = const[1]
>     MAD temp[0].xy, src0.xy_, src1.xy_, src0.000
>  8: BEGIN_TEX;
>  9: TEX temp[0], temp[0].xy__, 2D[0];
> 10: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[0]
>     MAD temp[1].xyz, src0.xyz, src1.yyy, src0.000
>     MAD temp[1].w, src0.w, src1.y, src0.0
> 11: BEGIN_TEX;
> 12: TEX temp[0], temp[2].xy__, 2D[0];
> 13: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[0], src1.w = temp[1], src2.xyz = temp[1]
>     MAD temp[1].xyz, src0.xyz, src1.xxx, src2.xyz
>     MAD temp[1].w, src0.w, src1.x, src1.w
> 14: BEGIN_TEX;
> 15: TEX temp[0], temp[3].xy__, 2D[0];
> 16: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[0], src1.w = temp[1], src2.xyz = temp[1]
>     MAD temp[1].xyz, src0.xyz, src1.zzz, src2.xyz
>     MAD temp[1].w, src0.w, src1.z, src1.w
> 17: BEGIN_TEX;
> 18: TEX temp[0], temp[4].xy__, 2D[0];
> 19: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = temp[1], src1.w = const[0], src2.w = temp[1]
>     MAD temp[1].xyz, src0.xyz, src1.www, src1.xyz
>     MAD temp[1].w, src0.w, src1.w, src2.w
> 20: BEGIN_TEX;
> 21: TEX temp[0], temp[5].xy__, 2D[0];
> 22: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = temp[1], src1.w = const[0], src2.w = temp[1]
>     MAD temp[1].xyz, src0.xyz, src1.www, src1.xyz
>     MAD temp[1].w, src0.w, src1.w, src2.w
> 23: BEGIN_TEX;
> 24: TEX temp[0], temp[6].xy__, 2D[0];
> 25: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[0], src1.w = temp[1], src2.xyz = temp[1]
>     MAD temp[1].xyz, src0.xyz, src1.zzz, src2.xyz
>     MAD temp[1].w, src0.w, src1.z, src1.w
> 26: BEGIN_TEX;
> 27: TEX temp[0], temp[7].xy__, 2D[0];
> 28: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[0], src1.w = temp[1], src2.xyz = temp[1]
>     MAD temp[1].xyz, src0.xyz, src1.yyy, src2.xyz
>     MAD temp[1].w, src0.w, src1.y, src1.w
> 29: BEGIN_TEX;
> 30: TEX temp[0], temp[8].xy__, 2D[0];
> 31: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[0], src1.w = temp[1], src2.xyz = temp[1]
>     MAD color[0].xyz, src0.xyz, src1.xxx, src2.xyz
>     MAD color[0].w, src0.w, src1.x, src1.w
>r300compiler error: r300_fragprog_emit.c::begin_tex(): Too many texture indirections
>pc=16*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 0, tex_end: 0  (code_addr: 00000000)
>  0: xyz:  t7  c1  t0 -> t8.xy                (01a00847)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t7.xyz   c1.xyz      0.0    op: 00050200
>       w:     t7.x     t7.x     t7.x    op: 00000000
>r300 FP: Compiler Error:
>r300_fragprog_emit.c::begin_tex(): Too many texture indirections
>Using a dummy shader instead.
>If there's an 'unknown opcode' message, please file a bug report and attach this log.
>r300: Initial fragment program
>FRAG
>DCL OUT[0], COLOR
>IMM FLT32 {    0.0000,     1.0000,     0.0000,     0.0000}
>  0: MOV OUT[0], IMM[0].xxxy
>  1: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0001;
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0001;
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0001;
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0001;
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0001;
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0001;
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0001;
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0001;
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: 
>     MAD color[0].xyz, src0.000, src0.111, src0.000
>     MAD color[0].w, src0.1, src0.1, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: 
>     MAD color[0].xyz, src0.000, src0.111, src0.000
>     MAD color[0].w, src0.1, src0.1, src0.0
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: 
>     MAD color[0].xyz, src0.000, src0.111, src0.000
>     MAD color[0].w, src0.1, src0.1, src0.0
>pc=17*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 0, tex_end: 0  (code_addr: 00400000)
>  0: xyz:  t0  t0  t0 -> o0.xyz               (1c000000)
>       w:  t0  t0  t0 -> o0.w                 (01000000)
>     xyz:      0.0      1.0      0.0    op: 00050a94
>       w:      1.0      1.0      0.0    op: 00040891
>r300: Initial fragment program
>FRAG
>DCL IN[0], GENERIC[0], PERSPECTIVE
>DCL IN[1], GENERIC[1], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL SAMP[1]
>DCL CONST[0]
>DCL CONST[2..3]
>DCL TEMP[0..1]
>IMM FLT32 {    0.2990,     0.5870,     0.1140,     0.5000}
>  0: TEX TEMP[0], IN[1], SAMP[0], RECT
>  1: MUL TEMP[1], TEMP[0], CONST[0].yyyy
>  2: TEX TEMP[0], IN[0], SAMP[1], RECT
>  3: MAD TEMP[0], TEMP[0], CONST[0].xxxx, TEMP[1]
>  4: DP3 TEMP[1].x, TEMP[0], IMM[0]
>  5: ADD TEMP[0], TEMP[0], -TEMP[1].xxxx
>  6: MAD TEMP[0], TEMP[0], CONST[2].xxxx, TEMP[1].xxxx
>  7: ADD TEMP[0], TEMP[0], -IMM[0].wwww
>  8: MUL TEMP[0], TEMP[0], CONST[2].yyyy
>  9: ADD TEMP[0], TEMP[0], IMM[0].wwww
> 10: MUL OUT[0], TEMP[0], CONST[3]
> 11: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: TEX temp[0], input[1], RECT[0];
>  1: MUL temp[1], temp[0], const[0].yyyy;
>  2: TEX temp[0], input[0], RECT[1];
>  3: MAD temp[0], temp[0], const[0].xxxx, temp[1];
>  4: DP3 temp[1].x, temp[0], const[4];
>  5: ADD temp[0], temp[0], -temp[1].xxxx;
>  6: MAD temp[0], temp[0], const[2].xxxx, temp[1].xxxx;
>  7: ADD temp[0], temp[0], -const[4].wwww;
>  8: MUL temp[0], temp[0], const[2].yyyy;
>  9: ADD temp[0], temp[0], const[4].wwww;
> 10: MUL output[0], temp[0], const[3];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: TEX temp[0], input[1], RECT[0];
>  1: MUL temp[1], temp[0], const[0].yyyy;
>  2: TEX temp[0], input[0], RECT[1];
>  3: MAD temp[0], temp[0], const[0].xxxx, temp[1];
>  4: DP3 temp[1].x, temp[0], const[4];
>  5: ADD temp[0], temp[0], -temp[1].xxxx;
>  6: MAD temp[0], temp[0], const[2].xxxx, temp[1].xxxx;
>  7: ADD temp[0], temp[0], -const[4].wwww;
>  8: MUL temp[0], temp[0], const[2].yyyy;
>  9: ADD temp[0], temp[0], const[4].wwww;
> 10: MUL output[0], temp[0], const[3];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: TEX temp[0], input[1], RECT[0];
>  1: MUL temp[1], temp[0], const[0].yyyy;
>  2: TEX temp[0], input[0], RECT[1];
>  3: MAD temp[0], temp[0], const[0].xxxx, temp[1];
>  4: DP3 temp[1].x, temp[0], const[4];
>  5: ADD temp[0], temp[0], -temp[1].xxxx;
>  6: MAD temp[0], temp[0], const[2].xxxx, temp[1].xxxx;
>  7: ADD temp[0], temp[0], -const[4].wwww;
>  8: MUL temp[0], temp[0], const[2].yyyy;
>  9: ADD temp[0], temp[0], const[4].wwww;
> 10: MUL output[0], temp[0], const[3];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: TEX temp[0], input[1], RECT[0];
>  1: MUL temp[1], temp[0], const[0].yyyy;
>  2: TEX temp[0], input[0], RECT[1];
>  3: MAD temp[0], temp[0], const[0].xxxx, temp[1];
>  4: DP3 temp[1].x, temp[0], const[4];
>  5: ADD temp[0], temp[0], -temp[1].xxxx;
>  6: MAD temp[0], temp[0], const[2].xxxx, temp[1].xxxx;
>  7: ADD temp[0], temp[0], -const[4].wwww;
>  8: MUL temp[0], temp[0], const[2].yyyy;
>  9: ADD temp[0], temp[0], const[4].wwww;
> 10: MUL output[0], temp[0], const[3];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: MUL temp[2], input[1], const[5];
>  1: TEX temp[0], temp[2], 2D[0];
>  2: MUL temp[1], temp[0], const[0].yyyy;
>  3: MUL temp[3], input[0], const[6];
>  4: TEX temp[0], temp[3], 2D[1];
>  5: MAD temp[0], temp[0], const[0].xxxx, temp[1];
>  6: DP3 temp[1].x, temp[0], const[4];
>  7: ADD temp[0], temp[0], -temp[1].xxxx;
>  8: MAD temp[0], temp[0], const[2].xxxx, temp[1].xxxx;
>  9: ADD temp[0], temp[0], -const[4].wwww;
> 10: MUL temp[0], temp[0], const[2].yyyy;
> 11: ADD temp[0], temp[0], const[4].wwww;
> 12: MUL output[0], temp[0], const[3];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: MUL temp[2].xy, input[1].xy__, const[5].xy__;
>  1: TEX temp[0], temp[2].xy__, 2D[0];
>  2: MUL temp[1], temp[0], const[0].yyyy;
>  3: MUL temp[3].xy, input[0].xy__, const[6].xy__;
>  4: TEX temp[0], temp[3].xy__, 2D[1];
>  5: MAD temp[0], temp[0], const[0].xxxx, temp[1];
>  6: DP3 temp[1].x, temp[0].xyz_, const[4].xyz_;
>  7: ADD temp[0], temp[0], -temp[1].xxxx;
>  8: MAD temp[0], temp[0], const[2].xxxx, temp[1].xxxx;
>  9: ADD temp[0], temp[0], -const[4].wwww;
> 10: MUL temp[0], temp[0], const[2].yyyy;
> 11: ADD temp[0], temp[0], const[4].wwww;
> 12: MUL output[0], temp[0], const[3];
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: MUL temp[2].xy, input[1].xy__, const[5].xy__;
>  1: TEX temp[0], temp[2].xy__, 2D[0];
>  2: MUL temp[1], temp[0], const[0].yyyy;
>  3: MUL temp[3].xy, input[0].xy__, const[6].xy__;
>  4: TEX temp[0], temp[3].xy__, 2D[1];
>  5: MAD temp[0], temp[0], const[0].xxxx, temp[1];
>  6: DP3 temp[1].x, temp[0].xyz_, const[4].xyz_;
>  7: ADD temp[0], temp[0], -temp[1].xxxx;
>  8: MAD temp[0], temp[0], const[2].xxxx, temp[1].xxxx;
>  9: ADD temp[0], temp[0], -none.HHHH;
> 10: MUL temp[0], temp[0], const[2].yyyy;
> 11: ADD temp[0], temp[0], none.HHHH;
> 12: MUL output[0], temp[0], const[3];
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: MUL temp[2].xy, input[1].xy__, const[5].xy__;
>  1: TEX temp[0], temp[2].xy__, 2D[0];
>  2: MUL temp[1], temp[0], const[0].yyyy;
>  3: MUL temp[3].xy, input[0].xy__, const[6].xy__;
>  4: TEX temp[0], temp[3].xy__, 2D[1];
>  5: MAD temp[0], temp[0], const[0].xxxx, temp[1];
>  6: DP3 temp[1].x, temp[0].xyz_, const[4].xyz_;
>  7: ADD temp[0], temp[0], -temp[1].xxxx;
>  8: MAD temp[0], temp[0], const[2].xxxx, temp[1].xxxx;
>  9: ADD temp[0], temp[0], -none.HHHH;
> 10: MUL temp[0], temp[0], const[2].yyyy;
> 11: ADD temp[0], temp[0], none.HHHH;
> 12: MUL output[0], temp[0], const[3];
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: src0.xyz = input[1], src1.xyz = const[5]
>     MAD temp[2].xy, src0.xy_, src1.xy_, src0.000
>  1: TEX temp[0], temp[2].xy__, 2D[0];
>  2: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[0]
>     MAD temp[1].xyz, src0.xyz, src1.yyy, src0.000
>     MAD temp[1].w, src0.w, src1.y, src0.0
>  3: src0.xyz = input[0], src1.xyz = const[6]
>     MAD temp[3].xy, src0.xy_, src1.xy_, src0.000
>  4: TEX temp[0], temp[3].xy__, 2D[1];
>  5: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[0], src1.w = temp[1], src2.xyz = temp[1]
>     MAD temp[0].xyz, src0.xyz, src1.xxx, src2.xyz
>     MAD temp[0].w, src0.w, src1.x, src1.w
>  6: src0.xyz = temp[0], src1.xyz = const[4]
>     DP3 temp[1].x, src0.xyz, src1.xyz
>  7: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = temp[1]
>     MAD temp[0].xyz, src0.xyz, src0.111, -src1.xxx
>     MAD temp[0].w, src0.w, src0.1, -src1.x
>  8: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[2], src2.xyz = temp[1]
>     MAD temp[0].xyz, src0.xyz, src1.xxx, src2.xxx
>     MAD temp[0].w, src0.w, src1.x, src2.x
>  9: src0.xyz = temp[0], src0.w = temp[0]
>     MAD temp[0].xyz, src0.xyz, src0.111, -src0.HHH
>     MAD temp[0].w, src0.w, src0.1, -src0.H
> 10: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[2]
>     MAD temp[0].xyz, src0.xyz, src1.yyy, src0.000
>     MAD temp[0].w, src0.w, src1.y, src0.0
> 11: src0.xyz = temp[0], src0.w = temp[0]
>     MAD temp[0].xyz, src0.xyz, src0.111, src0.HHH
>     MAD temp[0].w, src0.w, src0.1, src0.H
> 12: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[3], src1.w = const[3]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: src0.xyz = input[0], src1.xyz = const[6]
>     MAD temp[3].xy, src0.xy_, src1.xy_, src0.000
>  1: src0.xyz = input[1], src1.xyz = const[5]
>     MAD temp[2].xy, src0.xy_, src1.xy_, src0.000
>  2: BEGIN_TEX;
>  3: TEX temp[0], temp[2].xy__, 2D[0];
>  4: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[0]
>     MAD temp[1].xyz, src0.xyz, src1.yyy, src0.000
>     MAD temp[1].w, src0.w, src1.y, src0.0
>  5: BEGIN_TEX;
>  6: TEX temp[0], temp[3].xy__, 2D[1];
>  7: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[0], src1.w = temp[1], src2.xyz = temp[1]
>     MAD temp[0].xyz, src0.xyz, src1.xxx, src2.xyz
>     MAD temp[0].w, src0.w, src1.x, src1.w
>  8: src0.xyz = temp[0], src1.xyz = const[4]
>     DP3 temp[1].x, src0.xyz, src1.xyz
>  9: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = temp[1]
>     MAD temp[0].xyz, src0.xyz, src0.111, -src1.xxx
>     MAD temp[0].w, src0.w, src0.1, -src1.x
> 10: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[2], src2.xyz = temp[1]
>     MAD temp[0].xyz, src0.xyz, src1.xxx, src2.xxx
>     MAD temp[0].w, src0.w, src1.x, src2.x
> 11: src0.xyz = temp[0], src0.w = temp[0]
>     MAD temp[0].xyz, src0.xyz, src0.111, -src0.HHH
>     MAD temp[0].w, src0.w, src0.1, -src0.H
> 12: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[2]
>     MAD temp[0].xyz, src0.xyz, src1.yyy, src0.000
>     MAD temp[0].w, src0.w, src1.y, src0.0
> 13: src0.xyz = temp[0], src0.w = temp[0]
>     MAD temp[0].xyz, src0.xyz, src0.111, src0.HHH
>     MAD temp[0].w, src0.w, src0.1, src0.H
> 14: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[3], src1.w = const[3]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: src0.xyz = temp[0], src1.xyz = const[6]
>     MAD temp[2].xy, src0.xy_, src1.xy_, src0.000
>  1: src0.xyz = temp[1], src1.xyz = const[5]
>     MAD temp[0].xy, src0.xy_, src1.xy_, src0.000
>  2: BEGIN_TEX;
>  3: TEX temp[0], temp[0].xy__, 2D[0];
>  4: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[0]
>     MAD temp[1].xyz, src0.xyz, src1.yyy, src0.000
>     MAD temp[1].w, src0.w, src1.y, src0.0
>  5: BEGIN_TEX;
>  6: TEX temp[0], temp[2].xy__, 2D[1];
>  7: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[0], src1.w = temp[1], src2.xyz = temp[1]
>     MAD temp[0].xyz, src0.xyz, src1.xxx, src2.xyz
>     MAD temp[0].w, src0.w, src1.x, src1.w
>  8: src0.xyz = temp[0], src1.xyz = const[4]
>     DP3 temp[1].x, src0.xyz, src1.xyz
>  9: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = temp[1]
>     MAD temp[0].xyz, src0.xyz, src0.111, -src1.xxx
>     MAD temp[0].w, src0.w, src0.1, -src1.x
> 10: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[2], src2.xyz = temp[1]
>     MAD temp[0].xyz, src0.xyz, src1.xxx, src2.xxx
>     MAD temp[0].w, src0.w, src1.x, src2.x
> 11: src0.xyz = temp[0], src0.w = temp[0]
>     MAD temp[0].xyz, src0.xyz, src0.111, -src0.HHH
>     MAD temp[0].w, src0.w, src0.1, -src0.H
> 12: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[2]
>     MAD temp[0].xyz, src0.xyz, src1.yyy, src0.000
>     MAD temp[0].w, src0.w, src1.y, src0.0
> 13: src0.xyz = temp[0], src0.w = temp[0]
>     MAD temp[0].xyz, src0.xyz, src0.111, src0.HHH
>     MAD temp[0].w, src0.w, src0.1, src0.H
> 14: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[3], src1.w = const[3]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>pc=18*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 1, tex_end: 0  (code_addr: 00000040)
>  0: xyz:  t0  c6  t0 -> t2.xy                (01880980)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xyz   c6.xyz      0.0    op: 00050200
>       w:     t0.x     t0.x     t0.x    op: 00000000
>  1: xyz:  t1  c5  t0 -> t0.xy                (01800941)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   c5.xyz      0.0    op: 00050200
>       w:     t1.x     t1.x     t1.x    op: 00000000
>NODE 1: alu_offset: 2, tex_offset: 0, alu_end: 0, tex_end: 0  (code_addr: 00000002)
>  TEX:
>    TEX t0, t0, texture[0]   (00008000)
>  2: xyz:  t0  c0  t0 -> t1.xyz               (03840800)
>       w:  t0  t0  t0 -> t1.w                 (00840000)
>     xyz:   t0.xyz   c0.yyy      0.0    op: 00050300
>       w:     t0.w     c0.y      0.0    op: 00040209
>NODE 2: alu_offset: 3, tex_offset: 1, alu_end: 7, tex_end: 0  (code_addr: 004011c3)
>  TEX:
>    TEX t0, t2, texture[1]   (00008802)
>  3: xyz:  t0  c0  t1 -> t0.xyz               (03801800)
>       w:  t0  t1  t0 -> t0.w                 (00800040)
>     xyz:   t0.xyz   c0.xxx   t1.xyz    op: 00020280
>       w:     t0.w     c0.x     t1.w    op: 00028189
>  4: xyz:  t0  c4  t0 -> t1.x                 (00840900)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xyz   c4.xyz   t0.xxx    op: 00804200
>       w:     t0.x     t0.x     t0.x    op: 00000000
>  5: xyz:  t0  t1  t0 -> t0.xyz               (03800040)
>       w:  t0  t0  t0 -> t0.w                 (00800000)
>     xyz:   t0.xyz      1.0  -t1.xxx    op: 00094a80
>       w:     t0.w      1.0    -t1.x    op: 0008c889
>  6: xyz:  t0  c2  t1 -> t0.xyz               (03801880)
>       w:  t0  t0  t0 -> t0.w                 (00800000)
>     xyz:   t0.xyz   c2.xxx   t1.xxx    op: 00024280
>       w:     t0.w     c2.x     t1.x    op: 00018189
>  7: xyz:  t0  t0  t0 -> t0.xyz               (03800000)
>       w:  t0  t0  t0 -> t0.w                 (00800000)
>     xyz:   t0.xyz      1.0     -0.5    op: 000d8a80
>       w:     t0.w      1.0     -0.5    op: 000c8889
>  8: xyz:  t0  c2  t0 -> t0.xyz               (03800880)
>       w:  t0  t0  t0 -> t0.w                 (00800000)
>     xyz:   t0.xyz   c2.yyy      0.0    op: 00050300
>       w:     t0.w     c2.y      0.0    op: 00040209
>  9: xyz:  t0  t0  t0 -> t0.xyz               (03800000)
>       w:  t0  t0  t0 -> t0.w                 (00800000)
>     xyz:   t0.xyz      1.0      0.5    op: 00058a80
>       w:     t0.w      1.0      0.5    op: 00048889
> 10: xyz:  t0  c3  t0 -> o0.xyz               (1c0008c0)
>       w:  t0  c3  t0 -> o0.w                 (010008c0)
>     xyz:   t0.xyz   c3.xyz      0.0    op: 00050200
>       w:     t0.w     c3.w      0.0    op: 00040509
>r300: Initial fragment program
>FRAG
>DCL IN[0], GENERIC[0], PERSPECTIVE
>DCL IN[1], GENERIC[1], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL SAMP[1]
>DCL CONST[0]
>DCL CONST[2..3]
>DCL TEMP[0..1]
>IMM FLT32 {    0.2990,     0.5870,     0.1140,     0.5000}
>  0: TEX TEMP[0], IN[1], SAMP[0], RECT
>  1: MUL TEMP[1], TEMP[0], CONST[0].yyyy
>  2: TEX TEMP[0], IN[0], SAMP[1], RECT
>  3: MAD TEMP[0], TEMP[0], CONST[0].xxxx, TEMP[1]
>  4: DP3 TEMP[1].x, TEMP[0], IMM[0]
>  5: ADD TEMP[0], TEMP[0], -TEMP[1].xxxx
>  6: MAD TEMP[0], TEMP[0], CONST[2].xxxx, TEMP[1].xxxx
>  7: ADD TEMP[0], TEMP[0], -IMM[0].wwww
>  8: MUL TEMP[0], TEMP[0], CONST[2].yyyy
>  9: ADD TEMP[0], TEMP[0], IMM[0].wwww
> 10: MUL OUT[0], TEMP[0], CONST[3]
> 11: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: TEX temp[0], input[1], RECT[0];
>  1: MUL temp[1], temp[0], const[0].yyyy;
>  2: TEX temp[0], input[0], RECT[1];
>  3: MAD temp[0], temp[0], const[0].xxxx, temp[1];
>  4: DP3 temp[1].x, temp[0], const[4];
>  5: ADD temp[0], temp[0], -temp[1].xxxx;
>  6: MAD temp[0], temp[0], const[2].xxxx, temp[1].xxxx;
>  7: ADD temp[0], temp[0], -const[4].wwww;
>  8: MUL temp[0], temp[0], const[2].yyyy;
>  9: ADD temp[0], temp[0], const[4].wwww;
> 10: MUL output[0], temp[0], const[3];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: TEX temp[0], input[1], RECT[0];
>  1: MUL temp[1], temp[0], const[0].yyyy;
>  2: TEX temp[0], input[0], RECT[1];
>  3: MAD temp[0], temp[0], const[0].xxxx, temp[1];
>  4: DP3 temp[1].x, temp[0], const[4];
>  5: ADD temp[0], temp[0], -temp[1].xxxx;
>  6: MAD temp[0], temp[0], const[2].xxxx, temp[1].xxxx;
>  7: ADD temp[0], temp[0], -const[4].wwww;
>  8: MUL temp[0], temp[0], const[2].yyyy;
>  9: ADD temp[0], temp[0], const[4].wwww;
> 10: MUL output[0], temp[0], const[3];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: TEX temp[0], input[1], RECT[0];
>  1: MUL temp[1], temp[0], const[0].yyyy;
>  2: TEX temp[0], input[0], RECT[1];
>  3: MAD temp[0], temp[0], const[0].xxxx, temp[1];
>  4: DP3 temp[1].x, temp[0], const[4];
>  5: ADD temp[0], temp[0], -temp[1].xxxx;
>  6: MAD temp[0], temp[0], const[2].xxxx, temp[1].xxxx;
>  7: ADD temp[0], temp[0], -const[4].wwww;
>  8: MUL temp[0], temp[0], const[2].yyyy;
>  9: ADD temp[0], temp[0], const[4].wwww;
> 10: MUL output[0], temp[0], const[3];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: TEX temp[0], input[1], RECT[0];
>  1: MUL temp[1], temp[0], const[0].yyyy;
>  2: TEX temp[0], input[0], RECT[1];
>  3: MAD temp[0], temp[0], const[0].xxxx, temp[1];
>  4: DP3 temp[1].x, temp[0], const[4];
>  5: ADD temp[0], temp[0], -temp[1].xxxx;
>  6: MAD temp[0], temp[0], const[2].xxxx, temp[1].xxxx;
>  7: ADD temp[0], temp[0], -const[4].wwww;
>  8: MUL temp[0], temp[0], const[2].yyyy;
>  9: ADD temp[0], temp[0], const[4].wwww;
> 10: MUL output[0], temp[0], const[3];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: MUL temp[2], input[1], const[5];
>  1: TEX temp[0], temp[2], 2D[0];
>  2: MUL temp[1], temp[0], const[0].yyyy;
>  3: MUL temp[3], input[0], const[6];
>  4: TEX temp[0], temp[3], 2D[1];
>  5: MAD temp[0], temp[0], const[0].xxxx, temp[1];
>  6: DP3 temp[1].x, temp[0], const[4];
>  7: ADD temp[0], temp[0], -temp[1].xxxx;
>  8: MAD temp[0], temp[0], const[2].xxxx, temp[1].xxxx;
>  9: ADD temp[0], temp[0], -const[4].wwww;
> 10: MUL temp[0], temp[0], const[2].yyyy;
> 11: ADD temp[0], temp[0], const[4].wwww;
> 12: MUL output[0], temp[0], const[3];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: MUL temp[2].xy, input[1].xy__, const[5].xy__;
>  1: TEX temp[0], temp[2].xy__, 2D[0];
>  2: MUL temp[1], temp[0], const[0].yyyy;
>  3: MUL temp[3].xy, input[0].xy__, const[6].xy__;
>  4: TEX temp[0], temp[3].xy__, 2D[1];
>  5: MAD temp[0], temp[0], const[0].xxxx, temp[1];
>  6: DP3 temp[1].x, temp[0].xyz_, const[4].xyz_;
>  7: ADD temp[0], temp[0], -temp[1].xxxx;
>  8: MAD temp[0], temp[0], const[2].xxxx, temp[1].xxxx;
>  9: ADD temp[0], temp[0], -const[4].wwww;
> 10: MUL temp[0], temp[0], const[2].yyyy;
> 11: ADD temp[0], temp[0], const[4].wwww;
> 12: MUL output[0], temp[0], const[3];
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: MUL temp[2].xy, input[1].xy__, const[5].xy__;
>  1: TEX temp[0], temp[2].xy__, 2D[0];
>  2: MUL temp[1], temp[0], const[0].yyyy;
>  3: MUL temp[3].xy, input[0].xy__, const[6].xy__;
>  4: TEX temp[0], temp[3].xy__, 2D[1];
>  5: MAD temp[0], temp[0], const[0].xxxx, temp[1];
>  6: DP3 temp[1].x, temp[0].xyz_, const[4].xyz_;
>  7: ADD temp[0], temp[0], -temp[1].xxxx;
>  8: MAD temp[0], temp[0], const[2].xxxx, temp[1].xxxx;
>  9: ADD temp[0], temp[0], -none.HHHH;
> 10: MUL temp[0], temp[0], const[2].yyyy;
> 11: ADD temp[0], temp[0], none.HHHH;
> 12: MUL output[0], temp[0], const[3];
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: MUL temp[2].xy, input[1].xy__, const[5].xy__;
>  1: TEX temp[0], temp[2].xy__, 2D[0];
>  2: MUL temp[1], temp[0], const[0].yyyy;
>  3: MUL temp[3].xy, input[0].xy__, const[6].xy__;
>  4: TEX temp[0], temp[3].xy__, 2D[1];
>  5: MAD temp[0], temp[0], const[0].xxxx, temp[1];
>  6: DP3 temp[1].x, temp[0].xyz_, const[4].xyz_;
>  7: ADD temp[0], temp[0], -temp[1].xxxx;
>  8: MAD temp[0], temp[0], const[2].xxxx, temp[1].xxxx;
>  9: ADD temp[0], temp[0], -none.HHHH;
> 10: MUL temp[0], temp[0], const[2].yyyy;
> 11: ADD temp[0], temp[0], none.HHHH;
> 12: MUL output[0], temp[0], const[3];
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: src0.xyz = input[1], src1.xyz = const[5]
>     MAD temp[2].xy, src0.xy_, src1.xy_, src0.000
>  1: TEX temp[0], temp[2].xy__, 2D[0];
>  2: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[0]
>     MAD temp[1].xyz, src0.xyz, src1.yyy, src0.000
>     MAD temp[1].w, src0.w, src1.y, src0.0
>  3: src0.xyz = input[0], src1.xyz = const[6]
>     MAD temp[3].xy, src0.xy_, src1.xy_, src0.000
>  4: TEX temp[0], temp[3].xy__, 2D[1];
>  5: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[0], src1.w = temp[1], src2.xyz = temp[1]
>     MAD temp[0].xyz, src0.xyz, src1.xxx, src2.xyz
>     MAD temp[0].w, src0.w, src1.x, src1.w
>  6: src0.xyz = temp[0], src1.xyz = const[4]
>     DP3 temp[1].x, src0.xyz, src1.xyz
>  7: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = temp[1]
>     MAD temp[0].xyz, src0.xyz, src0.111, -src1.xxx
>     MAD temp[0].w, src0.w, src0.1, -src1.x
>  8: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[2], src2.xyz = temp[1]
>     MAD temp[0].xyz, src0.xyz, src1.xxx, src2.xxx
>     MAD temp[0].w, src0.w, src1.x, src2.x
>  9: src0.xyz = temp[0], src0.w = temp[0]
>     MAD temp[0].xyz, src0.xyz, src0.111, -src0.HHH
>     MAD temp[0].w, src0.w, src0.1, -src0.H
> 10: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[2]
>     MAD temp[0].xyz, src0.xyz, src1.yyy, src0.000
>     MAD temp[0].w, src0.w, src1.y, src0.0
> 11: src0.xyz = temp[0], src0.w = temp[0]
>     MAD temp[0].xyz, src0.xyz, src0.111, src0.HHH
>     MAD temp[0].w, src0.w, src0.1, src0.H
> 12: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[3], src1.w = const[3]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: src0.xyz = input[0], src1.xyz = const[6]
>     MAD temp[3].xy, src0.xy_, src1.xy_, src0.000
>  1: src0.xyz = input[1], src1.xyz = const[5]
>     MAD temp[2].xy, src0.xy_, src1.xy_, src0.000
>  2: BEGIN_TEX;
>  3: TEX temp[0], temp[2].xy__, 2D[0];
>  4: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[0]
>     MAD temp[1].xyz, src0.xyz, src1.yyy, src0.000
>     MAD temp[1].w, src0.w, src1.y, src0.0
>  5: BEGIN_TEX;
>  6: TEX temp[0], temp[3].xy__, 2D[1];
>  7: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[0], src1.w = temp[1], src2.xyz = temp[1]
>     MAD temp[0].xyz, src0.xyz, src1.xxx, src2.xyz
>     MAD temp[0].w, src0.w, src1.x, src1.w
>  8: src0.xyz = temp[0], src1.xyz = const[4]
>     DP3 temp[1].x, src0.xyz, src1.xyz
>  9: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = temp[1]
>     MAD temp[0].xyz, src0.xyz, src0.111, -src1.xxx
>     MAD temp[0].w, src0.w, src0.1, -src1.x
> 10: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[2], src2.xyz = temp[1]
>     MAD temp[0].xyz, src0.xyz, src1.xxx, src2.xxx
>     MAD temp[0].w, src0.w, src1.x, src2.x
> 11: src0.xyz = temp[0], src0.w = temp[0]
>     MAD temp[0].xyz, src0.xyz, src0.111, -src0.HHH
>     MAD temp[0].w, src0.w, src0.1, -src0.H
> 12: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[2]
>     MAD temp[0].xyz, src0.xyz, src1.yyy, src0.000
>     MAD temp[0].w, src0.w, src1.y, src0.0
> 13: src0.xyz = temp[0], src0.w = temp[0]
>     MAD temp[0].xyz, src0.xyz, src0.111, src0.HHH
>     MAD temp[0].w, src0.w, src0.1, src0.H
> 14: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[3], src1.w = const[3]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: src0.xyz = temp[0], src1.xyz = const[6]
>     MAD temp[2].xy, src0.xy_, src1.xy_, src0.000
>  1: src0.xyz = temp[1], src1.xyz = const[5]
>     MAD temp[0].xy, src0.xy_, src1.xy_, src0.000
>  2: BEGIN_TEX;
>  3: TEX temp[0], temp[0].xy__, 2D[0];
>  4: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[0]
>     MAD temp[1].xyz, src0.xyz, src1.yyy, src0.000
>     MAD temp[1].w, src0.w, src1.y, src0.0
>  5: BEGIN_TEX;
>  6: TEX temp[0], temp[2].xy__, 2D[1];
>  7: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[0], src1.w = temp[1], src2.xyz = temp[1]
>     MAD temp[0].xyz, src0.xyz, src1.xxx, src2.xyz
>     MAD temp[0].w, src0.w, src1.x, src1.w
>  8: src0.xyz = temp[0], src1.xyz = const[4]
>     DP3 temp[1].x, src0.xyz, src1.xyz
>  9: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = temp[1]
>     MAD temp[0].xyz, src0.xyz, src0.111, -src1.xxx
>     MAD temp[0].w, src0.w, src0.1, -src1.x
> 10: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[2], src2.xyz = temp[1]
>     MAD temp[0].xyz, src0.xyz, src1.xxx, src2.xxx
>     MAD temp[0].w, src0.w, src1.x, src2.x
> 11: src0.xyz = temp[0], src0.w = temp[0]
>     MAD temp[0].xyz, src0.xyz, src0.111, -src0.HHH
>     MAD temp[0].w, src0.w, src0.1, -src0.H
> 12: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[2]
>     MAD temp[0].xyz, src0.xyz, src1.yyy, src0.000
>     MAD temp[0].w, src0.w, src1.y, src0.0
> 13: src0.xyz = temp[0], src0.w = temp[0]
>     MAD temp[0].xyz, src0.xyz, src0.111, src0.HHH
>     MAD temp[0].w, src0.w, src0.1, src0.H
> 14: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[3], src1.w = const[3]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>pc=19*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 1, tex_end: 0  (code_addr: 00000040)
>  0: xyz:  t0  c6  t0 -> t2.xy                (01880980)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xyz   c6.xyz      0.0    op: 00050200
>       w:     t0.x     t0.x     t0.x    op: 00000000
>  1: xyz:  t1  c5  t0 -> t0.xy                (01800941)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   c5.xyz      0.0    op: 00050200
>       w:     t1.x     t1.x     t1.x    op: 00000000
>NODE 1: alu_offset: 2, tex_offset: 0, alu_end: 0, tex_end: 0  (code_addr: 00000002)
>  TEX:
>    TEX t0, t0, texture[0]   (00008000)
>  2: xyz:  t0  c0  t0 -> t1.xyz               (03840800)
>       w:  t0  t0  t0 -> t1.w                 (00840000)
>     xyz:   t0.xyz   c0.yyy      0.0    op: 00050300
>       w:     t0.w     c0.y      0.0    op: 00040209
>NODE 2: alu_offset: 3, tex_offset: 1, alu_end: 7, tex_end: 0  (code_addr: 004011c3)
>  TEX:
>    TEX t0, t2, texture[1]   (00008802)
>  3: xyz:  t0  c0  t1 -> t0.xyz               (03801800)
>       w:  t0  t1  t0 -> t0.w                 (00800040)
>     xyz:   t0.xyz   c0.xxx   t1.xyz    op: 00020280
>       w:     t0.w     c0.x     t1.w    op: 00028189
>  4: xyz:  t0  c4  t0 -> t1.x                 (00840900)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xyz   c4.xyz   t0.xxx    op: 00804200
>       w:     t0.x     t0.x     t0.x    op: 00000000
>  5: xyz:  t0  t1  t0 -> t0.xyz               (03800040)
>       w:  t0  t0  t0 -> t0.w                 (00800000)
>     xyz:   t0.xyz      1.0  -t1.xxx    op: 00094a80
>       w:     t0.w      1.0    -t1.x    op: 0008c889
>  6: xyz:  t0  c2  t1 -> t0.xyz               (03801880)
>       w:  t0  t0  t0 -> t0.w                 (00800000)
>     xyz:   t0.xyz   c2.xxx   t1.xxx    op: 00024280
>       w:     t0.w     c2.x     t1.x    op: 00018189
>  7: xyz:  t0  t0  t0 -> t0.xyz               (03800000)
>       w:  t0  t0  t0 -> t0.w                 (00800000)
>     xyz:   t0.xyz      1.0     -0.5    op: 000d8a80
>       w:     t0.w      1.0     -0.5    op: 000c8889
>  8: xyz:  t0  c2  t0 -> t0.xyz               (03800880)
>       w:  t0  t0  t0 -> t0.w                 (00800000)
>     xyz:   t0.xyz   c2.yyy      0.0    op: 00050300
>       w:     t0.w     c2.y      0.0    op: 00040209
>  9: xyz:  t0  t0  t0 -> t0.xyz               (03800000)
>       w:  t0  t0  t0 -> t0.w                 (00800000)
>     xyz:   t0.xyz      1.0      0.5    op: 00058a80
>       w:     t0.w      1.0      0.5    op: 00048889
> 10: xyz:  t0  c3  t0 -> o0.xyz               (1c0008c0)
>       w:  t0  c3  t0 -> o0.w                 (010008c0)
>     xyz:   t0.xyz   c3.xyz      0.0    op: 00050200
>       w:     t0.w     c3.w      0.0    op: 00040509
>r300: Initial fragment program
>FRAG
>DCL IN[0], COLOR, LINEAR
>DCL IN[1], GENERIC[0], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL TEMP[0]
>  0: TEX TEMP[0], IN[1], SAMP[0], 2D
>  1: MUL OUT[0], TEMP[0], IN[0]
>  2: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: MUL output[0], temp[0], input[0];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: MUL output[0], temp[0], input[0];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: MUL output[0], temp[0], input[0];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: MUL output[0], temp[0], input[0];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: MUL temp[1], input[1], const[0];
>  1: TEX temp[0], temp[1], 2D[0];
>  2: MUL output[0], temp[0], input[0];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: MUL temp[1].xy, input[1].xy__, const[0].xy__;
>  1: TEX temp[0], temp[1].xy__, 2D[0];
>  2: MUL output[0], temp[0], input[0];
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: MUL temp[1].xy, input[1].xy__, const[0].xy__;
>  1: TEX temp[0], temp[1].xy__, 2D[0];
>  2: MUL output[0], temp[0], input[0];
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: MUL temp[1].xy, input[1].xy__, const[0].xy__;
>  1: TEX temp[0], temp[1].xy__, 2D[0];
>  2: MUL output[0], temp[0], input[0];
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: src0.xyz = input[1], src1.xyz = const[0]
>     MAD temp[1].xy, src0.xy_, src1.xy_, src0.000
>  1: TEX temp[0], temp[1].xy__, 2D[0];
>  2: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = input[0], src1.w = input[0]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: src0.xyz = input[1], src1.xyz = const[0]
>     MAD temp[1].xy, src0.xy_, src1.xy_, src0.000
>  1: BEGIN_TEX;
>  2: TEX temp[0], temp[1].xy__, 2D[0];
>  3: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = input[0], src1.w = input[0]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: src0.xyz = temp[1], src1.xyz = const[0]
>     MAD temp[1].xy, src0.xy_, src1.xy_, src0.000
>  1: BEGIN_TEX;
>  2: TEX temp[1], temp[1].xy__, 2D[0];
>  3: src0.xyz = temp[1], src0.w = temp[1], src1.xyz = temp[0], src1.w = temp[0]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>pc=20*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 0, tex_end: 0  (code_addr: 00000000)
>  0: xyz:  t1  c0  t0 -> t1.xy                (01840801)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   c0.xyz      0.0    op: 00050200
>       w:     t1.x     t1.x     t1.x    op: 00000000
>NODE 1: alu_offset: 1, tex_offset: 0, alu_end: 0, tex_end: 0  (code_addr: 00400001)
>  TEX:
>    TEX t1, t1, texture[0]   (00008041)
>  1: xyz:  t1  t0  t0 -> o0.xyz               (1c000001)
>       w:  t1  t0  t0 -> o0.w                 (01000001)
>     xyz:   t1.xyz   t0.xyz      0.0    op: 00050200
>       w:     t1.w     t0.w      0.0    op: 00040509
>r300: Initial fragment program
>FRAG
>DCL IN[0], COLOR, LINEAR
>DCL IN[1], GENERIC[0], PERSPECTIVE
>DCL IN[2], GENERIC[1], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL SAMP[1]
>DCL TEMP[0..1]
>  0: TEX TEMP[0], IN[1], SAMP[0], 2D
>  1: TEX TEMP[1], IN[2], SAMP[1], 2D
>  2: MUL TEMP[0], TEMP[0], TEMP[1]
>  3: MUL OUT[0], TEMP[0], IN[0]
>  4: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: TEX temp[1], input[2], 2D[1];
>  2: MUL temp[0], temp[0], temp[1];
>  3: MUL output[0], temp[0], input[0];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: TEX temp[1], input[2], 2D[1];
>  2: MUL temp[0], temp[0], temp[1];
>  3: MUL output[0], temp[0], input[0];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: TEX temp[1], input[2], 2D[1];
>  2: MUL temp[0], temp[0], temp[1];
>  3: MUL output[0], temp[0], input[0];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: TEX temp[1], input[2], 2D[1];
>  2: MUL temp[0], temp[0], temp[1];
>  3: MUL output[0], temp[0], input[0];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: TEX temp[1], input[2], 2D[1];
>  2: MUL temp[0], temp[0], temp[1];
>  3: MUL output[0], temp[0], input[0];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: TEX temp[0], input[1].xy__, 2D[0];
>  1: TEX temp[1], input[2].xy__, 2D[1];
>  2: MUL temp[0], temp[0], temp[1];
>  3: MUL output[0], temp[0], input[0];
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: TEX temp[0], input[1].xy__, 2D[0];
>  1: TEX temp[1], input[2].xy__, 2D[1];
>  2: MUL temp[0], temp[0], temp[1];
>  3: MUL output[0], temp[0], input[0];
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: TEX temp[0], input[1].xy__, 2D[0];
>  1: TEX temp[1], input[2].xy__, 2D[1];
>  2: MUL temp[0], temp[0], temp[1];
>  3: MUL output[0], temp[0], input[0];
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: TEX temp[0], input[1].xy__, 2D[0];
>  1: TEX temp[1], input[2].xy__, 2D[1];
>  2: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = temp[1], src1.w = temp[1]
>     MAD temp[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD temp[0].w, src0.w, src1.w, src0.0
>  3: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = input[0], src1.w = input[0]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[1], input[2].xy__, 2D[1];
>  2: TEX temp[0], input[1].xy__, 2D[0];
>  3: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = temp[1], src1.w = temp[1]
>     MAD temp[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD temp[0].w, src0.w, src1.w, src0.0
>  4: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = input[0], src1.w = input[0]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[2], temp[2].xy__, 2D[1];
>  2: TEX temp[1], temp[1].xy__, 2D[0];
>  3: src0.xyz = temp[1], src0.w = temp[1], src1.xyz = temp[2], src1.w = temp[2]
>     MAD temp[1].xyz, src0.xyz, src1.xyz, src0.000
>     MAD temp[1].w, src0.w, src1.w, src0.0
>  4: src0.xyz = temp[1], src0.w = temp[1], src1.xyz = temp[0], src1.w = temp[0]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>pc=21*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 1, tex_end: 1  (code_addr: 00420040)
>  TEX:
>    TEX t2, t2, texture[1]   (00008882)
>    TEX t1, t1, texture[0]   (00008041)
>  0: xyz:  t1  t2  t0 -> t1.xyz               (03840081)
>       w:  t1  t2  t0 -> t1.w                 (00840081)
>     xyz:   t1.xyz   t2.xyz      0.0    op: 00050200
>       w:     t1.w     t2.w      0.0    op: 00040509
>  1: xyz:  t1  t0  t0 -> o0.xyz               (1c000001)
>       w:  t1  t0  t0 -> o0.w                 (01000001)
>     xyz:   t1.xyz   t0.xyz      0.0    op: 00050200
>       w:     t1.w     t0.w      0.0    op: 00040509
>r300: Initial fragment program
>FRAG
>DCL IN[0], COLOR, LINEAR
>DCL IN[1], GENERIC[0], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL TEMP[0]
>IMM FLT32 {    1.0000,     0.0000,     0.0000,     0.0000}
>  0: TEX TEMP[0], IN[1], SAMP[0], 2D
>  1: SUB TEMP[0], TEMP[0], IMM[0].xxxx
>  2: MAD OUT[0].xyz, TEMP[0], IN[0].wwww, IMM[0].xxxx
>  3: MOV OUT[0].w, IMM[0].xxxx
>  4: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: SUB temp[0], temp[0], temp[0].1111;
>  2: MAD output[0].xyz, temp[0], input[0].wwww, temp[0].1111;
>  3: MOV output[0].w, temp[0].1111;
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: SUB temp[0], temp[0], temp[0].1111;
>  2: MAD output[0].xyz, temp[0], input[0].wwww, temp[0].1111;
>  3: MOV output[0].w, temp[0].1111;
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: SUB temp[0], temp[0], temp[0].1111;
>  2: MAD output[0].xyz, temp[0], input[0].wwww, temp[0].1111;
>  3: MOV output[0].w, temp[0].1111;
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: ADD temp[0], temp[0], -temp[0].1111;
>  2: MAD output[0].xyz, temp[0], input[0].wwww, temp[0].1111;
>  3: MOV output[0].w, temp[0].1111;
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: ADD temp[0], temp[0], -temp[0].1111;
>  2: MAD output[0].xyz, temp[0], input[0].wwww, temp[0].1111;
>  3: MOV output[0].w, temp[0].1111;
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: TEX temp[0].xyz, input[1].xy__, 2D[0];
>  1: ADD temp[0].xyz, temp[0].xyz_, -temp[0].111_;
>  2: MAD output[0].xyz, temp[0].xyz_, input[0].www_, temp[0].111_;
>  3: MOV output[0].w, temp[0].___1;
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: TEX temp[0].xyz, input[1].xy__, 2D[0];
>  1: ADD temp[0].xyz, temp[0].xyz_, -temp[0].111_;
>  2: MAD output[0].xyz, temp[0].xyz_, input[0].www_, temp[0].111_;
>  3: MOV output[0].w, temp[0].___1;
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: TEX temp[0].xyz, input[1].xy__, 2D[0];
>  1: ADD temp[0].xyz, temp[0].xyz_, -temp[0].111_;
>  2: MAD output[0].xyz, temp[0].xyz_, input[0].www_, temp[0].111_;
>  3: MOV output[0].w, temp[0].___1;
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: TEX temp[0].xyz, input[1].xy__, 2D[0];
>  1: src0.xyz = temp[0]
>     MAD temp[0].xyz, src0.xyz, src0.111, -src0.111
>  2: src0.xyz = temp[0], src0.w = input[0]
>     MAD color[0].xyz, src0.xyz, src0.www, src0.111
>  3: 
>     MAD color[0].w, src0.1, src0.1, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[0].xyz, input[1].xy__, 2D[0];
>  2: src0.xyz = temp[0]
>     MAD temp[0].xyz, src0.xyz, src0.111, -src0.111
>     MAD color[0].w, src0.1, src0.1, src0.0
>  3: src0.xyz = temp[0], src0.w = input[0]
>     MAD color[0].xyz, src0.xyz, src0.www, src0.111
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[1].xyz, temp[1].xy__, 2D[0];
>  2: src0.xyz = temp[1]
>     MAD temp[1].xyz, src0.xyz, src0.111, -src0.111
>     MAD color[0].w, src0.1, src0.1, src0.0
>  3: src0.xyz = temp[1], src0.w = temp[0]
>     MAD color[0].xyz, src0.xyz, src0.www, src0.111
>pc=22*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 1, tex_end: 0  (code_addr: 00400040)
>  TEX:
>    TEX t1, t1, texture[0]   (00008041)
>  0: xyz:  t1  t0  t0 -> t1.xyz               (03840001)
>       w:  t0  t0  t0 -> o0.w                 (01000000)
>     xyz:   t1.xyz      1.0     -1.0    op: 000d4a80
>       w:      1.0      1.0      0.0    op: 00040891
>  1: xyz:  t1  t0  t0 -> o0.xyz               (1c000001)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   t0.www      1.0    op: 00054600
>       w:     t1.x     t1.x     t1.x    op: 00000000
>r300: Initial fragment program
>FRAG
>DCL IN[0], COLOR, LINEAR
>DCL IN[1], GENERIC[0], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL TEMP[0]
>IMM FLT32 {    1.0000,     0.0000,     0.0000,     0.0000}
>  0: TEX TEMP[0].xyz, IN[1], SAMP[0], 2D
>  1: MOV TEMP[0].w, IMM[0].xxxx
>  2: MUL OUT[0], TEMP[0], IN[0]
>  3: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: TEX temp[0].xyz, input[1], 2D[0];
>  1: MOV temp[0].w, temp[0].1111;
>  2: MUL output[0], temp[0], input[0];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: TEX temp[0].xyz, input[1], 2D[0];
>  1: MOV temp[0].w, temp[0].1111;
>  2: MUL output[0], temp[0], input[0];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: TEX temp[0].xyz, input[1], 2D[0];
>  1: MOV temp[0].w, temp[0].1111;
>  2: MUL output[0], temp[0], input[0];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: TEX temp[0].xyz, input[1], 2D[0];
>  1: MOV temp[0].w, temp[0].1111;
>  2: MUL output[0], temp[0], input[0];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: TEX temp[1], input[1], 2D[0];
>  1: MOV temp[0].xyz, temp[1];
>  2: MOV temp[0].w, temp[0].1111;
>  3: MUL output[0], temp[0], input[0];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: TEX temp[1].xyz, input[1].xy__, 2D[0];
>  1: MOV temp[0].xyz, temp[1].xyz_;
>  2: MOV temp[0].w, temp[0].___1;
>  3: MUL output[0], temp[0], input[0];
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: TEX temp[1].xyz, input[1].xy__, 2D[0];
>  1: MOV temp[0].xyz, temp[1].xyz_;
>  2: MOV temp[0].w, temp[0].___1;
>  3: MUL output[0], temp[0], input[0];
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: TEX temp[1].xyz, input[1].xy__, 2D[0];
>  1: MOV temp[0].xyz, temp[1].xyz_;
>  2: MOV temp[0].w, temp[0].___1;
>  3: MUL output[0], temp[0], input[0];
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: TEX temp[1].xyz, input[1].xy__, 2D[0];
>  1: src0.xyz = temp[1]
>     MAD temp[0].xyz, src0.xyz, src0.111, src0.000
>  2: 
>     MAD temp[0].w, src0.1, src0.1, src0.0
>  3: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = input[0], src1.w = input[0]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[1].xyz, input[1].xy__, 2D[0];
>  2: src0.xyz = temp[1]
>     MAD temp[0].xyz, src0.xyz, src0.111, src0.000
>     MAD temp[0].w, src0.1, src0.1, src0.0
>  3: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = input[0], src1.w = input[0]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[1].xyz, temp[1].xy__, 2D[0];
>  2: src0.xyz = temp[1]
>     MAD temp[1].xyz, src0.xyz, src0.111, src0.000
>     MAD temp[1].w, src0.1, src0.1, src0.0
>  3: src0.xyz = temp[1], src0.w = temp[1], src1.xyz = temp[0], src1.w = temp[0]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>pc=23*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 1, tex_end: 0  (code_addr: 00400040)
>  TEX:
>    TEX t1, t1, texture[0]   (00008041)
>  0: xyz:  t1  t0  t0 -> t1.xyz               (03840001)
>       w:  t0  t0  t0 -> t1.w                 (00840000)
>     xyz:   t1.xyz      1.0      0.0    op: 00050a80
>       w:      1.0      1.0      0.0    op: 00040891
>  1: xyz:  t1  t0  t0 -> o0.xyz               (1c000001)
>       w:  t1  t0  t0 -> o0.w                 (01000001)
>     xyz:   t1.xyz   t0.xyz      0.0    op: 00050200
>       w:     t1.w     t0.w      0.0    op: 00040509
>r300: Initial fragment program
>FRAG
>DCL IN[0], POSITION, LINEAR
>DCL IN[1], COLOR, LINEAR
>DCL IN[2], COLOR[1], LINEAR
>DCL IN[3], GENERIC[2], PERSPECTIVE
>DCL IN[4], GENERIC[3], PERSPECTIVE
>DCL IN[5], GENERIC[4], PERSPECTIVE
>DCL IN[6], GENERIC[6], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL SAMP[1]
>DCL SAMP[2]
>DCL SAMP[3]
>DCL CONST[4..5]
>DCL CONST[1..3]
>DCL TEMP[0..4]
>IMM FLT32 {    0.0625,     1.0000,     0.0000,     0.0000}
>  0: MOV TEMP[0], IN[0]
>  1: SUB TEMP[0].y, CONST[5], IN[0]
>  2: TXP TEMP[1], IN[4], SAMP[0], 2D
>  3: MUL TEMP[2].xy, TEMP[0], IMM[0].xxxx
>  4: TEX TEMP[2], TEMP[2], SAMP[1], 2D
>  5: SLT TEMP[2].w, IN[6].wwww, TEMP[2].wwww
>  6: KIL -TEMP[2].wwww
>  7: TEX TEMP[2], IN[3], SAMP[2], 2D
>  8: MUL TEMP[1], TEMP[1], TEMP[2]
>  9: TEX TEMP[2], IN[5], SAMP[3], 2D
> 10: MUL TEMP[2].xyz, TEMP[2], IN[1]
> 11: MUL TEMP[3], TEMP[2], CONST[1]
> 12: SUB TEMP[4].w, TEMP[2].wwww, CONST[2]
> 13: KIL TEMP[4].wwww
> 14: MUL TEMP[3], TEMP[1], TEMP[3]
> 15: MAD TEMP[3], TEMP[2], CONST[3], TEMP[3]
> 16: LRP OUT[0].xyz, IN[2].xxxx, CONST[4], TEMP[3]
> 17: MOV OUT[0].w, IMM[0].yyyy
> 18: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: RCP temp[5].w, input[0].wwww;
>  1: MUL temp[5].xyz, input[0], temp[5].wwww;
>  2: MAD temp[5].xyz, temp[5].xyz0, const[7].xyz0, const[8].xyz0;
>  3: MOV temp[0], temp[5];
>  4: SUB temp[0].y, const[5], temp[5];
>  5: TXP temp[1], input[4], 2D[0];
>  6: MUL temp[2].xy, temp[0], const[6].xxxx;
>  7: TEX temp[2], temp[2], 2D[1];
>  8: SLT temp[2].w, input[6].wwww, temp[2].wwww;
>  9: KIL -temp[2].wwww;
> 10: TEX temp[2], input[3], 2D[2];
> 11: MUL temp[1], temp[1], temp[2];
> 12: TEX temp[2], input[5], 2D[3];
> 13: MUL temp[2].xyz, temp[2], input[1];
> 14: MUL temp[3], temp[2], const[1];
> 15: SUB temp[4].w, temp[2].wwww, const[2];
> 16: KIL temp[4].wwww;
> 17: MUL temp[3], temp[1], temp[3];
> 18: MAD temp[3], temp[2], const[3], temp[3];
> 19: LRP output[0].xyz, input[2].xxxx, const[4], temp[3];
> 20: MOV output[0].w, const[6].yyyy;
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: RCP temp[5].w, input[0].wwww;
>  1: MUL temp[5].xyz, input[0], temp[5].wwww;
>  2: MAD temp[5].xyz, temp[5].xyz0, const[7].xyz0, const[8].xyz0;
>  3: MOV temp[0], temp[5];
>  4: SUB temp[0].y, const[5], temp[5];
>  5: TXP temp[1], input[4], 2D[0];
>  6: MUL temp[2].xy, temp[0], const[6].xxxx;
>  7: TEX temp[2], temp[2], 2D[1];
>  8: SLT temp[2].w, input[6].wwww, temp[2].wwww;
>  9: KIL -temp[2].wwww;
> 10: TEX temp[2], input[3], 2D[2];
> 11: MUL temp[1], temp[1], temp[2];
> 12: TEX temp[2], input[5], 2D[3];
> 13: MUL temp[2].xyz, temp[2], input[1];
> 14: MUL temp[3], temp[2], const[1];
> 15: SUB temp[4].w, temp[2].wwww, const[2];
> 16: KIL temp[4].wwww;
> 17: MUL temp[3], temp[1], temp[3];
> 18: MAD temp[3], temp[2], const[3], temp[3];
> 19: LRP output[0].xyz, input[2].xxxx, const[4], temp[3];
> 20: MOV output[0].w, const[6].yyyy;
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: RCP temp[5].w, input[0].wwww;
>  1: MUL temp[5].xyz, input[0], temp[5].wwww;
>  2: MAD temp[5].xyz, temp[5].xyz0, const[7].xyz0, const[8].xyz0;
>  3: MOV temp[0], temp[5];
>  4: SUB temp[0].y, const[5], temp[5];
>  5: TXP temp[1], input[4], 2D[0];
>  6: MUL temp[2].xy, temp[0], const[6].xxxx;
>  7: TEX temp[2], temp[2], 2D[1];
>  8: SLT temp[2].w, input[6].wwww, temp[2].wwww;
>  9: KIL -temp[2].wwww;
> 10: TEX temp[2], input[3], 2D[2];
> 11: MUL temp[1], temp[1], temp[2];
> 12: TEX temp[2], input[5], 2D[3];
> 13: MUL temp[2].xyz, temp[2], input[1];
> 14: MUL temp[3], temp[2], const[1];
> 15: SUB temp[4].w, temp[2].wwww, const[2];
> 16: KIL temp[4].wwww;
> 17: MUL temp[3], temp[1], temp[3];
> 18: MAD temp[3], temp[2], const[3], temp[3];
> 19: LRP output[0].xyz, input[2].xxxx, const[4], temp[3];
> 20: MOV output[0].w, const[6].yyyy;
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: RCP temp[5].w, input[0].wwww;
>  1: MUL temp[5].xyz, input[0], temp[5].wwww;
>  2: MAD temp[5].xyz, temp[5].xyz0, const[7].xyz0, const[8].xyz0;
>  3: MOV temp[0], temp[5];
>  4: ADD temp[0].y, const[5], -temp[5];
>  5: TXP temp[1], input[4], 2D[0];
>  6: MUL temp[2].xy, temp[0], const[6].xxxx;
>  7: TEX temp[2], temp[2], 2D[1];
>  8: ADD temp[6], input[6].wwww, -temp[2].wwww;
>  9: CMP temp[2].w, temp[6], none.1111, none.0000;
> 10: KIL -temp[2].wwww;
> 11: TEX temp[2], input[3], 2D[2];
> 12: MUL temp[1], temp[1], temp[2];
> 13: TEX temp[2], input[5], 2D[3];
> 14: MUL temp[2].xyz, temp[2], input[1];
> 15: MUL temp[3], temp[2], const[1];
> 16: ADD temp[4].w, temp[2].wwww, -const[2];
> 17: KIL temp[4].wwww;
> 18: MUL temp[3], temp[1], temp[3];
> 19: MAD temp[3], temp[2], const[3], temp[3];
> 20: ADD temp[7], const[4], -temp[3];
> 21: MAD output[0].xyz, input[2].xxxx, temp[7], temp[3];
> 22: MOV output[0].w, const[6].yyyy;
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: RCP temp[5].w, input[0].wwww;
>  1: MUL temp[5].xyz, input[0], temp[5].wwww;
>  2: MAD temp[5].xyz, temp[5].xyz0, const[7].xyz0, const[8].xyz0;
>  3: MOV temp[0], temp[5];
>  4: ADD temp[0].y, const[5], -temp[5];
>  5: TXP temp[1], input[4], 2D[0];
>  6: MUL temp[2].xy, temp[0], const[6].xxxx;
>  7: TEX temp[2], temp[2], 2D[1];
>  8: ADD temp[6], input[6].wwww, -temp[2].wwww;
>  9: CMP temp[2].w, temp[6], none.1111, none.0000;
> 10: KIL -temp[2].wwww;
> 11: TEX temp[2], input[3], 2D[2];
> 12: MUL temp[1], temp[1], temp[2];
> 13: TEX temp[2], input[5], 2D[3];
> 14: MUL temp[2].xyz, temp[2], input[1];
> 15: MUL temp[3], temp[2], const[1];
> 16: ADD temp[4].w, temp[2].wwww, -const[2];
> 17: KIL temp[4].wwww;
> 18: MUL temp[3], temp[1], temp[3];
> 19: MAD temp[3], temp[2], const[3], temp[3];
> 20: ADD temp[7], const[4], -temp[3];
> 21: MAD output[0].xyz, input[2].xxxx, temp[7], temp[3];
> 22: MOV output[0].w, const[6].yyyy;
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: RCP temp[5].w, input[0].w___;
>  1: MUL temp[5].xy, input[0].xy__, temp[5].ww__;
>  2: MAD temp[5].xy, temp[5].xy__, const[7].xy__, const[8].xy__;
>  3: MOV temp[0].x, temp[5].x___;
>  4: ADD temp[0].y, const[5]._y__, -temp[5]._y__;
>  5: TXP temp[1].xyz, input[4].xy_w, 2D[0];
>  6: MUL temp[2].xy, temp[0].xy__, const[6].xx__;
>  7: TEX temp[2].w, temp[2].xy__, 2D[1];
>  8: ADD temp[6].w, input[6].___w, -temp[2].___w;
>  9: CMP temp[2].w, temp[6].___w, none.___1, none.___0;
> 10: KIL -temp[2].wwww;
> 11: TEX temp[2].xyz, input[3].xy__, 2D[2];
> 12: MUL temp[1].xyz, temp[1].xyz_, temp[2].xyz_;
> 13: TEX temp[2], input[5].xy__, 2D[3];
> 14: MUL temp[2].xyz, temp[2].xyz_, input[1].xyz_;
> 15: MUL temp[3].xyz, temp[2].xyz_, const[1].xyz_;
> 16: ADD temp[4].w, temp[2].___w, -const[2].___w;
> 17: KIL temp[4].wwww;
> 18: MUL temp[3].xyz, temp[1].xyz_, temp[3].xyz_;
> 19: MAD temp[3].xyz, temp[2].xyz_, const[3].xyz_, temp[3].xyz_;
> 20: ADD temp[7].xyz, const[4].xyz_, -temp[3].xyz_;
> 21: MAD output[0].xyz, input[2].xxx_, temp[7].xyz_, temp[3].xyz_;
> 22: MOV output[0].w, const[6].___y;
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: RCP temp[5].w, input[0].w___;
>  1: MUL temp[5].xy, input[0].xy__, temp[5].ww__;
>  2: MAD temp[5].xy, temp[5].xy__, const[7].xy__, const[8].xy__;
>  3: MOV temp[0].x, temp[5].x___;
>  4: ADD temp[0].y, const[5]._y__, -temp[5]._y__;
>  5: TXP temp[1].xyz, input[4].xy_w, 2D[0];
>  6: MUL temp[2].xy, temp[0].xy__, const[6].xx__;
>  7: TEX temp[2].w, temp[2].xy__, 2D[1];
>  8: ADD temp[6].w, input[6].___w, -temp[2].___w;
>  9: CMP temp[2].w, temp[6].___w, none.___1, none.___0;
> 10: KIL -temp[2].wwww;
> 11: TEX temp[2].xyz, input[3].xy__, 2D[2];
> 12: MUL temp[1].xyz, temp[1].xyz_, temp[2].xyz_;
> 13: TEX temp[2], input[5].xy__, 2D[3];
> 14: MUL temp[2].xyz, temp[2].xyz_, input[1].xyz_;
> 15: MUL temp[3].xyz, temp[2].xyz_, const[1].xyz_;
> 16: ADD temp[4].w, temp[2].___w, -const[2].___w;
> 17: KIL temp[4].wwww;
> 18: MUL temp[3].xyz, temp[1].xyz_, temp[3].xyz_;
> 19: MAD temp[3].xyz, temp[2].xyz_, const[3].xyz_, temp[3].xyz_;
> 20: ADD temp[7].xyz, const[4].xyz_, -temp[3].xyz_;
> 21: MAD output[0].xyz, input[2].xxx_, temp[7].xyz_, temp[3].xyz_;
> 22: MOV output[0].w, none.___1;
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: RCP temp[5].w, input[0].w___;
>  1: MUL temp[5].xy, input[0].xy__, temp[5].ww__;
>  2: MAD temp[5].xy, temp[5].xy__, const[7].xy__, const[8].xy__;
>  3: MOV temp[0].x, temp[5].x___;
>  4: ADD temp[0].y, const[5]._y__, -temp[5]._y__;
>  5: TXP temp[1].xyz, input[4].xy_w, 2D[0];
>  6: MUL temp[2].xy, temp[0].xy__, const[6].xx__;
>  7: TEX temp[2].w, temp[2].xy__, 2D[1];
>  8: ADD temp[6].w, input[6].___w, -temp[2].___w;
>  9: CMP temp[2].w, temp[6].___w, none.___1, none.___0;
> 10: MOV temp[8], -temp[2].wwww;
> 11: KIL temp[8];
> 12: TEX temp[2].xyz, input[3].xy__, 2D[2];
> 13: MUL temp[1].xyz, temp[1].xyz_, temp[2].xyz_;
> 14: TEX temp[2], input[5].xy__, 2D[3];
> 15: MUL temp[2].xyz, temp[2].xyz_, input[1].xyz_;
> 16: MUL temp[3].xyz, temp[2].xyz_, const[1].xyz_;
> 17: ADD temp[4].w, temp[2].___w, -const[2].___w;
> 18: MOV temp[9], temp[4].wwww;
> 19: KIL temp[9];
> 20: MUL temp[3].xyz, temp[1].xyz_, temp[3].xyz_;
> 21: MAD temp[3].xyz, temp[2].xyz_, const[3].xyz_, temp[3].xyz_;
> 22: ADD temp[7].xyz, const[4].xyz_, -temp[3].xyz_;
> 23: MAD output[0].xyz, input[2].xxx_, temp[7].xyz_, temp[3].xyz_;
> 24: MOV output[0].w, none.___1;
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: src0.w = input[0]
>     RCP temp[5].w, src0.w
>  1: src0.xyz = input[0], src0.w = temp[5]
>     MAD temp[5].xy, src0.xy_, src0.ww_, src0.000
>  2: src0.xyz = temp[5], src1.xyz = const[7], src2.xyz = const[8]
>     MAD temp[5].xy, src0.xy_, src1.xy_, src2.xy_
>  3: src0.xyz = temp[5]
>     MAD temp[0].x, src0.x__, src0.111, src0.000
>  4: src0.xyz = const[5], src1.xyz = temp[5]
>     MAD temp[0].y, src0._y_, src0.111, -src1._y_
>  5: TXP temp[1].xyz, input[4].xy_w, 2D[0];
>  6: src0.xyz = temp[0], src1.xyz = const[6]
>     MAD temp[2].xy, src0.xy_, src1.xx_, src0.000
>  7: TEX temp[2].w, temp[2].xy__, 2D[1];
>  8: src0.w = input[6], src1.w = temp[2]
>     MAD temp[6].w, src0.w, src0.1, -src1.w
>  9: src0.w = temp[6]
>     CMP temp[2].w, src0.0, src0.1, src0.w
> 10: src0.w = temp[2]
>     MAD temp[8].xyz, -src0.www, src0.111, src0.000
>     MAD temp[8].w, -src0.w, src0.1, src0.0
> 11: KIL temp[8];
> 12: TEX temp[2].xyz, input[3].xy__, 2D[2];
> 13: src0.xyz = temp[1], src1.xyz = temp[2]
>     MAD temp[1].xyz, src0.xyz, src1.xyz, src0.000
> 14: TEX temp[2], input[5].xy__, 2D[3];
> 15: src0.xyz = temp[2], src1.xyz = input[1]
>     MAD temp[2].xyz, src0.xyz, src1.xyz, src0.000
> 16: src0.xyz = temp[2], src1.xyz = const[1]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src0.000
> 17: src0.w = temp[2], src1.w = const[2]
>     MAD temp[4].w, src0.w, src0.1, -src1.w
> 18: src0.w = temp[4]
>     MAD temp[9].xyz, src0.www, src0.111, src0.000
>     MAD temp[9].w, src0.w, src0.1, src0.0
> 19: KIL temp[9];
> 20: src0.xyz = temp[1], src1.xyz = temp[3]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src0.000
> 21: src0.xyz = temp[2], src1.xyz = const[3], src2.xyz = temp[3]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src2.xyz
> 22: src0.xyz = const[4], src1.xyz = temp[3]
>     MAD temp[7].xyz, src0.xyz, src0.111, -src1.xyz
> 23: src0.xyz = input[2], src1.xyz = temp[7], src2.xyz = temp[3]
>     MAD color[0].xyz, src0.xxx, src1.xyz, src2.xyz
> 24: 
>     MAD color[0].w, src0.1, src0.1, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TXP temp[1].xyz, input[4].xy_w, 2D[0];
>  2: 
>     MAD color[0].w, src0.1, src0.1, src0.0
>  3: src0.w = input[0]
>     RCP temp[5].w, src0.w
>  4: src0.xyz = input[0], src0.w = temp[5]
>     MAD temp[5].xy, src0.xy_, src0.ww_, src0.000
>  5: src0.xyz = temp[5], src1.xyz = const[7], src2.xyz = const[8]
>     MAD temp[5].xy, src0.xy_, src1.xy_, src2.xy_
>  6: src0.xyz = const[5], src1.xyz = temp[5]
>     MAD temp[0].y, src0._y_, src0.111, -src1._y_
>  7: src0.xyz = temp[5]
>     MAD temp[0].x, src0.x__, src0.111, src0.000
>  8: src0.xyz = temp[0], src1.xyz = const[6]
>     MAD temp[2].xy, src0.xy_, src1.xx_, src0.000
>  9: BEGIN_TEX;
> 10: TEX temp[2].w, temp[2].xy__, 2D[1];
> 11: src0.w = input[6], src1.w = temp[2]
>     MAD temp[6].w, src0.w, src0.1, -src1.w
> 12: src0.w = temp[6]
>     CMP temp[2].w, src0.0, src0.1, src0.w
> 13: src0.w = temp[2]
>     MAD temp[8].xyz, -src0.www, src0.111, src0.000
>     MAD temp[8].w, -src0.w, src0.1, src0.0
> 14: BEGIN_TEX;
> 15: KIL temp[8];
> 16: TEX temp[2].xyz, input[3].xy__, 2D[2];
> 17: src0.xyz = temp[1], src1.xyz = temp[2]
>     MAD temp[1].xyz, src0.xyz, src1.xyz, src0.000
> 18: BEGIN_TEX;
> 19: TEX temp[2], input[5].xy__, 2D[3];
> 20: src0.xyz = temp[2], src0.w = temp[2], src1.xyz = input[1], src1.w = const[2]
>     MAD temp[2].xyz, src0.xyz, src1.xyz, src0.000
>     MAD temp[4].w, src0.w, src0.1, -src1.w
> 21: src0.w = temp[4]
>     MAD temp[9].xyz, src0.www, src0.111, src0.000
>     MAD temp[9].w, src0.w, src0.1, src0.0
> 22: src0.xyz = temp[2], src1.xyz = const[1]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src0.000
> 23: src0.xyz = temp[1], src1.xyz = temp[3]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src0.000
> 24: src0.xyz = temp[2], src1.xyz = const[3], src2.xyz = temp[3]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src2.xyz
> 25: src0.xyz = const[4], src1.xyz = temp[3]
>     MAD temp[7].xyz, src0.xyz, src0.111, -src1.xyz
> 26: src0.xyz = input[2], src1.xyz = temp[7], src2.xyz = temp[3]
>     MAD color[0].xyz, src0.xxx, src1.xyz, src2.xyz
> 27: BEGIN_TEX;
> 28: KIL temp[9];
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TXP temp[7].xyz, temp[3].xy_w, 2D[0];
>  2: 
>     MAD color[0].w, src0.1, src0.1, src0.0
>  3: src0.w = temp[6]
>     RCP temp[8].w, src0.w
>  4: src0.xyz = temp[6], src0.w = temp[8]
>     MAD temp[8].xy, src0.xy_, src0.ww_, src0.000
>  5: src0.xyz = temp[8], src1.xyz = const[7], src2.xyz = const[8]
>     MAD temp[8].xy, src0.xy_, src1.xy_, src2.xy_
>  6: src0.xyz = const[5], src1.xyz = temp[8]
>     MAD temp[3].y, src0._y_, src0.111, -src1._y_
>  7: src0.xyz = temp[8]
>     MAD temp[3].x, src0.x__, src0.111, src0.000
>  8: src0.xyz = temp[3], src1.xyz = const[6]
>     MAD temp[3].xy, src0.xy_, src1.xx_, src0.000
>  9: BEGIN_TEX;
> 10: TEX temp[3].w, temp[3].xy__, 2D[1];
> 11: src0.w = temp[5], src1.w = temp[3]
>     MAD temp[5].w, src0.w, src0.1, -src1.w
> 12: src0.w = temp[5]
>     CMP temp[3].w, src0.0, src0.1, src0.w
> 13: src0.w = temp[3]
>     MAD temp[5].xyz, -src0.www, src0.111, src0.000
>     MAD temp[5].w, -src0.w, src0.1, src0.0
> 14: BEGIN_TEX;
> 15: KIL temp[5];
> 16: TEX temp[3].xyz, temp[2].xy__, 2D[2];
> 17: src0.xyz = temp[7], src1.xyz = temp[3]
>     MAD temp[7].xyz, src0.xyz, src1.xyz, src0.000
> 18: BEGIN_TEX;
> 19: TEX temp[3], temp[4].xy__, 2D[3];
> 20: src0.xyz = temp[3], src0.w = temp[3], src1.xyz = temp[0], src1.w = const[2]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src0.000
>     MAD temp[0].w, src0.w, src0.1, -src1.w
> 21: src0.w = temp[0]
>     MAD temp[4].xyz, src0.www, src0.111, src0.000
>     MAD temp[4].w, src0.w, src0.1, src0.0
> 22: src0.xyz = temp[3], src1.xyz = const[1]
>     MAD temp[0].xyz, src0.xyz, src1.xyz, src0.000
> 23: src0.xyz = temp[7], src1.xyz = temp[0]
>     MAD temp[0].xyz, src0.xyz, src1.xyz, src0.000
> 24: src0.xyz = temp[3], src1.xyz = const[3], src2.xyz = temp[0]
>     MAD temp[0].xyz, src0.xyz, src1.xyz, src2.xyz
> 25: src0.xyz = const[4], src1.xyz = temp[0]
>     MAD temp[2].xyz, src0.xyz, src0.111, -src1.xyz
> 26: src0.xyz = temp[1], src1.xyz = temp[2], src2.xyz = temp[0]
>     MAD color[0].xyz, src0.xxx, src1.xyz, src2.xyz
> 27: BEGIN_TEX;
> 28: KIL temp[4];
>r300compiler error: r300_fragprog_emit.c::begin_tex(): Too many texture indirections
>pc=24*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 0, tex_end: 0  (code_addr: 00000000)
>  TEX:
>    TXP t7, t3, texture[0]   (000181c3)
>  0: xyz:  t0  t0  t0 ->                      (00000000)
>       w:  t0  t0  t0 -> o0.w                 (01000000)
>     xyz:   t0.xxx   t0.xxx   t0.xxx    op: 00004081
>       w:      1.0      1.0      0.0    op: 00040891
>r300 FP: Compiler Error:
>r300_fragprog_emit.c::begin_tex(): Too many texture indirections
>Using a dummy shader instead.
>If there's an 'unknown opcode' message, please file a bug report and attach this log.
>r300: Initial fragment program
>FRAG
>DCL OUT[0], COLOR
>IMM FLT32 {    0.0000,     1.0000,     0.0000,     0.0000}
>  0: MOV OUT[0], IMM[0].xxxy
>  1: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0001;
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0001;
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0001;
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0001;
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0001;
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0001;
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0001;
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0001;
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: 
>     MAD color[0].xyz, src0.000, src0.111, src0.000
>     MAD color[0].w, src0.1, src0.1, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: 
>     MAD color[0].xyz, src0.000, src0.111, src0.000
>     MAD color[0].w, src0.1, src0.1, src0.0
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: 
>     MAD color[0].xyz, src0.000, src0.111, src0.000
>     MAD color[0].w, src0.1, src0.1, src0.0
>pc=25*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 0, tex_end: 0  (code_addr: 00400000)
>  0: xyz:  t0  t0  t0 -> o0.xyz               (1c000000)
>       w:  t0  t0  t0 -> o0.w                 (01000000)
>     xyz:      0.0      1.0      0.0    op: 00050a94
>       w:      1.0      1.0      0.0    op: 00040891
>r300: Initial fragment program
>FRAG
>DCL IN[0], POSITION, LINEAR
>DCL IN[1], COLOR, LINEAR
>DCL IN[2], COLOR[1], LINEAR
>DCL IN[3], GENERIC[0], PERSPECTIVE
>DCL IN[4], GENERIC[2], PERSPECTIVE
>DCL IN[5], GENERIC[3], PERSPECTIVE
>DCL IN[6], GENERIC[4], PERSPECTIVE
>DCL IN[7], GENERIC[6], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL SAMP[1]
>DCL SAMP[2]
>DCL SAMP[3]
>DCL CONST[5..6]
>DCL CONST[1]
>DCL CONST[3..4]
>DCL TEMP[0..7]
>IMM FLT32 {    0.0625,     0.3000,     0.5900,     0.1100}
>IMM FLT32 {    0.3000,     0.5900,     0.1100,     0.0000}
>  0: MOV TEMP[0], IN[0]
>  1: SUB TEMP[0].y, CONST[6], IN[0]
>  2: TEX TEMP[1], IN[6], SAMP[0], 2D
>  3: MUL TEMP[2].xy, TEMP[0], IMM[0].xxxx
>  4: TEX TEMP[2], TEMP[2], SAMP[1], 2D
>  5: SLT TEMP[2].w, IN[7].wwww, TEMP[2].wwww
>  6: KIL -TEMP[2].wwww
>  7: MOV TEMP[3], TEMP[2]
>  8: SUB TEMP[2].w, TEMP[1].wwww, CONST[1]
>  9: KIL TEMP[2].wwww
> 10: TXP TEMP[2], IN[5], SAMP[2], 2D
> 11: MUL TEMP[4], IN[3].zzzz, TEMP[2]
> 12: MAX TEMP[4], TEMP[4], IMM[1].wwww
> 13: TEX TEMP[2].x, IN[4], SAMP[3], 2D
> 14: MUL TEMP[4], TEMP[4], TEMP[2].xxxx
> 15: MUL TEMP[5], TEMP[1], IN[1]
> 16: MUL TEMP[6], TEMP[5], CONST[3]
> 17: MUL TEMP[7], TEMP[4], TEMP[6]
> 18: MAD TEMP[7], TEMP[5], CONST[4], TEMP[7]
> 19: MOV OUT[0].w, IMM[1].wwww
> 20: LRP OUT[0].xyz, IN[2].xxxx, CONST[5], TEMP[7]
> 21: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: RCP temp[8].w, input[0].wwww;
>  1: MUL temp[8].xyz, input[0], temp[8].wwww;
>  2: MAD temp[8].xyz, temp[8].xyz0, const[9].xyz0, const[10].xyz0;
>  3: MOV temp[0], temp[8];
>  4: SUB temp[0].y, const[6], temp[8];
>  5: TEX temp[1], input[6], 2D[0];
>  6: MUL temp[2].xy, temp[0], const[7].xxxx;
>  7: TEX temp[2], temp[2], 2D[1];
>  8: SLT temp[2].w, input[7].wwww, temp[2].wwww;
>  9: KIL -temp[2].wwww;
> 10: MOV temp[3], temp[2];
> 11: SUB temp[2].w, temp[1].wwww, const[1];
> 12: KIL temp[2].wwww;
> 13: TXP temp[2], input[5], 2D[2];
> 14: MUL temp[4], input[3].zzzz, temp[2];
> 15: MAX temp[4], temp[4], const[8].wwww;
> 16: TEX temp[2].x, input[4], 2D[3];
> 17: MUL temp[4], temp[4], temp[2].xxxx;
> 18: MUL temp[5], temp[1], input[1];
> 19: MUL temp[6], temp[5], const[3];
> 20: MUL temp[7], temp[4], temp[6];
> 21: MAD temp[7], temp[5], const[4], temp[7];
> 22: MOV output[0].w, const[8].wwww;
> 23: LRP output[0].xyz, input[2].xxxx, const[5], temp[7];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: RCP temp[8].w, input[0].wwww;
>  1: MUL temp[8].xyz, input[0], temp[8].wwww;
>  2: MAD temp[8].xyz, temp[8].xyz0, const[9].xyz0, const[10].xyz0;
>  3: MOV temp[0], temp[8];
>  4: SUB temp[0].y, const[6], temp[8];
>  5: TEX temp[1], input[6], 2D[0];
>  6: MUL temp[2].xy, temp[0], const[7].xxxx;
>  7: TEX temp[2], temp[2], 2D[1];
>  8: SLT temp[2].w, input[7].wwww, temp[2].wwww;
>  9: KIL -temp[2].wwww;
> 10: MOV temp[3], temp[2];
> 11: SUB temp[2].w, temp[1].wwww, const[1];
> 12: KIL temp[2].wwww;
> 13: TXP temp[2], input[5], 2D[2];
> 14: MUL temp[4], input[3].zzzz, temp[2];
> 15: MAX temp[4], temp[4], const[8].wwww;
> 16: TEX temp[2].x, input[4], 2D[3];
> 17: MUL temp[4], temp[4], temp[2].xxxx;
> 18: MUL temp[5], temp[1], input[1];
> 19: MUL temp[6], temp[5], const[3];
> 20: MUL temp[7], temp[4], temp[6];
> 21: MAD temp[7], temp[5], const[4], temp[7];
> 22: MOV output[0].w, const[8].wwww;
> 23: LRP output[0].xyz, input[2].xxxx, const[5], temp[7];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: RCP temp[8].w, input[0].wwww;
>  1: MUL temp[8].xyz, input[0], temp[8].wwww;
>  2: MAD temp[8].xyz, temp[8].xyz0, const[9].xyz0, const[10].xyz0;
>  3: MOV temp[0], temp[8];
>  4: SUB temp[0].y, const[6], temp[8];
>  5: TEX temp[1], input[6], 2D[0];
>  6: MUL temp[2].xy, temp[0], const[7].xxxx;
>  7: TEX temp[2], temp[2], 2D[1];
>  8: SLT temp[2].w, input[7].wwww, temp[2].wwww;
>  9: KIL -temp[2].wwww;
> 10: MOV temp[3], temp[2];
> 11: SUB temp[2].w, temp[1].wwww, const[1];
> 12: KIL temp[2].wwww;
> 13: TXP temp[2], input[5], 2D[2];
> 14: MUL temp[4], input[3].zzzz, temp[2];
> 15: MAX temp[4], temp[4], const[8].wwww;
> 16: TEX temp[2].x, input[4], 2D[3];
> 17: MUL temp[4], temp[4], temp[2].xxxx;
> 18: MUL temp[5], temp[1], input[1];
> 19: MUL temp[6], temp[5], const[3];
> 20: MUL temp[7], temp[4], temp[6];
> 21: MAD temp[7], temp[5], const[4], temp[7];
> 22: MOV output[0].w, const[8].wwww;
> 23: LRP output[0].xyz, input[2].xxxx, const[5], temp[7];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: RCP temp[8].w, input[0].wwww;
>  1: MUL temp[8].xyz, input[0], temp[8].wwww;
>  2: MAD temp[8].xyz, temp[8].xyz0, const[9].xyz0, const[10].xyz0;
>  3: MOV temp[0], temp[8];
>  4: ADD temp[0].y, const[6], -temp[8];
>  5: TEX temp[1], input[6], 2D[0];
>  6: MUL temp[2].xy, temp[0], const[7].xxxx;
>  7: TEX temp[2], temp[2], 2D[1];
>  8: ADD temp[9], input[7].wwww, -temp[2].wwww;
>  9: CMP temp[2].w, temp[9], none.1111, none.0000;
> 10: KIL -temp[2].wwww;
> 11: MOV temp[3], temp[2];
> 12: ADD temp[2].w, temp[1].wwww, -const[1];
> 13: KIL temp[2].wwww;
> 14: TXP temp[2], input[5], 2D[2];
> 15: MUL temp[4], input[3].zzzz, temp[2];
> 16: MAX temp[4], temp[4], const[8].wwww;
> 17: TEX temp[2].x, input[4], 2D[3];
> 18: MUL temp[4], temp[4], temp[2].xxxx;
> 19: MUL temp[5], temp[1], input[1];
> 20: MUL temp[6], temp[5], const[3];
> 21: MUL temp[7], temp[4], temp[6];
> 22: MAD temp[7], temp[5], const[4], temp[7];
> 23: MOV output[0].w, const[8].wwww;
> 24: ADD temp[10], const[5], -temp[7];
> 25: MAD output[0].xyz, input[2].xxxx, temp[10], temp[7];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: RCP temp[8].w, input[0].wwww;
>  1: MUL temp[8].xyz, input[0], temp[8].wwww;
>  2: MAD temp[8].xyz, temp[8].xyz0, const[9].xyz0, const[10].xyz0;
>  3: MOV temp[0], temp[8];
>  4: ADD temp[0].y, const[6], -temp[8];
>  5: TEX temp[1], input[6], 2D[0];
>  6: MUL temp[2].xy, temp[0], const[7].xxxx;
>  7: TEX temp[2], temp[2], 2D[1];
>  8: ADD temp[9], input[7].wwww, -temp[2].wwww;
>  9: CMP temp[2].w, temp[9], none.1111, none.0000;
> 10: KIL -temp[2].wwww;
> 11: MOV temp[3], temp[2];
> 12: ADD temp[2].w, temp[1].wwww, -const[1];
> 13: KIL temp[2].wwww;
> 14: TXP temp[2], input[5], 2D[2];
> 15: MUL temp[4], input[3].zzzz, temp[2];
> 16: MAX temp[4], temp[4], const[8].wwww;
> 17: TEX temp[11], input[4], 2D[3];
> 18: MOV temp[2].x, temp[11];
> 19: MUL temp[4], temp[4], temp[2].xxxx;
> 20: MUL temp[5], temp[1], input[1];
> 21: MUL temp[6], temp[5], const[3];
> 22: MUL temp[7], temp[4], temp[6];
> 23: MAD temp[7], temp[5], const[4], temp[7];
> 24: MOV output[0].w, const[8].wwww;
> 25: ADD temp[10], const[5], -temp[7];
> 26: MAD output[0].xyz, input[2].xxxx, temp[10], temp[7];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: RCP temp[8].w, input[0].w___;
>  1: MUL temp[8].xy, input[0].xy__, temp[8].ww__;
>  2: MAD temp[8].xy, temp[8].xy__, const[9].xy__, const[10].xy__;
>  3: MOV temp[0].x, temp[8].x___;
>  4: ADD temp[0].y, const[6]._y__, -temp[8]._y__;
>  5: TEX temp[1], input[6].xy__, 2D[0];
>  6: MUL temp[2].xy, temp[0].xy__, const[7].xx__;
>  7: TEX temp[2].w, temp[2].xy__, 2D[1];
>  8: ADD temp[9].w, input[7].___w, -temp[2].___w;
>  9: CMP temp[2].w, temp[9].___w, none.___1, none.___0;
> 10: KIL -temp[2].wwww;
> 11: ADD temp[2].w, temp[1].___w, -const[1].___w;
> 12: KIL temp[2].wwww;
> 13: TXP temp[2].xyz, input[5].xy_w, 2D[2];
> 14: MUL temp[4].xyz, input[3].zzz_, temp[2].xyz_;
> 15: MAX temp[4].xyz, temp[4].xyz_, const[8].www_;
> 16: TEX temp[11].x, input[4].xy__, 2D[3];
> 17: MOV temp[2].x, temp[11].x___;
> 18: MUL temp[4].xyz, temp[4].xyz_, temp[2].xxx_;
> 19: MUL temp[5].xyz, temp[1].xyz_, input[1].xyz_;
> 20: MUL temp[6].xyz, temp[5].xyz_, const[3].xyz_;
> 21: MUL temp[7].xyz, temp[4].xyz_, temp[6].xyz_;
> 22: MAD temp[7].xyz, temp[5].xyz_, const[4].xyz_, temp[7].xyz_;
> 23: MOV output[0].w, const[8].___w;
> 24: ADD temp[10].xyz, const[5].xyz_, -temp[7].xyz_;
> 25: MAD output[0].xyz, input[2].xxx_, temp[10].xyz_, temp[7].xyz_;
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: RCP temp[8].w, input[0].w___;
>  1: MUL temp[8].xy, input[0].xy__, temp[8].ww__;
>  2: MAD temp[8].xy, temp[8].xy__, const[9].xy__, const[10].xy__;
>  3: MOV temp[0].x, temp[8].x___;
>  4: ADD temp[0].y, const[6]._y__, -temp[8]._y__;
>  5: TEX temp[1], input[6].xy__, 2D[0];
>  6: MUL temp[2].xy, temp[0].xy__, const[7].xx__;
>  7: TEX temp[2].w, temp[2].xy__, 2D[1];
>  8: ADD temp[9].w, input[7].___w, -temp[2].___w;
>  9: CMP temp[2].w, temp[9].___w, none.___1, none.___0;
> 10: KIL -temp[2].wwww;
> 11: ADD temp[2].w, temp[1].___w, -const[1].___w;
> 12: KIL temp[2].wwww;
> 13: TXP temp[2].xyz, input[5].xy_w, 2D[2];
> 14: MUL temp[4].xyz, input[3].zzz_, temp[2].xyz_;
> 15: MAX temp[4].xyz, temp[4].xyz_, none.000_;
> 16: TEX temp[11].x, input[4].xy__, 2D[3];
> 17: MUL temp[4].xyz, temp[4].xyz_, temp[11].xxx_;
> 18: MUL temp[5].xyz, temp[1].xyz_, input[1].xyz_;
> 19: MUL temp[6].xyz, temp[5].xyz_, const[3].xyz_;
> 20: MUL temp[7].xyz, temp[4].xyz_, temp[6].xyz_;
> 21: MAD temp[7].xyz, temp[5].xyz_, const[4].xyz_, temp[7].xyz_;
> 22: MOV output[0].w, none.___0;
> 23: ADD temp[10].xyz, const[5].xyz_, -temp[7].xyz_;
> 24: MAD output[0].xyz, input[2].xxx_, temp[10].xyz_, temp[7].xyz_;
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: RCP temp[8].w, input[0].w___;
>  1: MUL temp[8].xy, input[0].xy__, temp[8].ww__;
>  2: MAD temp[8].xy, temp[8].xy__, const[9].xy__, const[10].xy__;
>  3: MOV temp[0].x, temp[8].x___;
>  4: ADD temp[0].y, const[6]._y__, -temp[8]._y__;
>  5: TEX temp[1], input[6].xy__, 2D[0];
>  6: MUL temp[2].xy, temp[0].xy__, const[7].xx__;
>  7: TEX temp[2].w, temp[2].xy__, 2D[1];
>  8: ADD temp[9].w, input[7].___w, -temp[2].___w;
>  9: CMP temp[2].w, temp[9].___w, none.___1, none.___0;
> 10: MOV temp[3], -temp[2].wwww;
> 11: KIL temp[3];
> 12: ADD temp[2].w, temp[1].___w, -const[1].___w;
> 13: MOV temp[12], temp[2].wwww;
> 14: KIL temp[12];
> 15: TXP temp[2].xyz, input[5].xy_w, 2D[2];
> 16: MUL temp[4].xyz, input[3].zzz_, temp[2].xyz_;
> 17: MAX temp[4].xyz, temp[4].xyz_, none.000_;
> 18: TEX temp[11].x, input[4].xy__, 2D[3];
> 19: MUL temp[4].xyz, temp[4].xyz_, temp[11].xxx_;
> 20: MUL temp[5].xyz, temp[1].xyz_, input[1].xyz_;
> 21: MUL temp[6].xyz, temp[5].xyz_, const[3].xyz_;
> 22: MUL temp[7].xyz, temp[4].xyz_, temp[6].xyz_;
> 23: MAD temp[7].xyz, temp[5].xyz_, const[4].xyz_, temp[7].xyz_;
> 24: MOV output[0].w, none.___0;
> 25: ADD temp[10].xyz, const[5].xyz_, -temp[7].xyz_;
> 26: MAD output[0].xyz, input[2].xxx_, temp[10].xyz_, temp[7].xyz_;
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: src0.w = input[0]
>     RCP temp[8].w, src0.w
>  1: src0.xyz = input[0], src0.w = temp[8]
>     MAD temp[8].xy, src0.xy_, src0.ww_, src0.000
>  2: src0.xyz = temp[8], src1.xyz = const[9], src2.xyz = const[10]
>     MAD temp[8].xy, src0.xy_, src1.xy_, src2.xy_
>  3: src0.xyz = temp[8]
>     MAD temp[0].x, src0.x__, src0.111, src0.000
>  4: src0.xyz = const[6], src1.xyz = temp[8]
>     MAD temp[0].y, src0._y_, src0.111, -src1._y_
>  5: TEX temp[1], input[6].xy__, 2D[0];
>  6: src0.xyz = temp[0], src1.xyz = const[7]
>     MAD temp[2].xy, src0.xy_, src1.xx_, src0.000
>  7: TEX temp[2].w, temp[2].xy__, 2D[1];
>  8: src0.w = input[7], src1.w = temp[2]
>     MAD temp[9].w, src0.w, src0.1, -src1.w
>  9: src0.w = temp[9]
>     CMP temp[2].w, src0.0, src0.1, src0.w
> 10: src0.w = temp[2]
>     MAD temp[3].xyz, -src0.www, src0.111, src0.000
>     MAD temp[3].w, -src0.w, src0.1, src0.0
> 11: KIL temp[3];
> 12: src0.w = temp[1], src1.w = const[1]
>     MAD temp[2].w, src0.w, src0.1, -src1.w
> 13: src0.w = temp[2]
>     MAD temp[12].xyz, src0.www, src0.111, src0.000
>     MAD temp[12].w, src0.w, src0.1, src0.0
> 14: KIL temp[12];
> 15: TXP temp[2].xyz, input[5].xy_w, 2D[2];
> 16: src0.xyz = input[3], src1.xyz = temp[2]
>     MAD temp[4].xyz, src0.zzz, src1.xyz, src0.000
> 17: src0.xyz = temp[4]
>     MAX temp[4].xyz, src0.xyz, src0.000
> 18: TEX temp[11].x, input[4].xy__, 2D[3];
> 19: src0.xyz = temp[4], src1.xyz = temp[11]
>     MAD temp[4].xyz, src0.xyz, src1.xxx, src0.000
> 20: src0.xyz = temp[1], src1.xyz = input[1]
>     MAD temp[5].xyz, src0.xyz, src1.xyz, src0.000
> 21: src0.xyz = temp[5], src1.xyz = const[3]
>     MAD temp[6].xyz, src0.xyz, src1.xyz, src0.000
> 22: src0.xyz = temp[4], src1.xyz = temp[6]
>     MAD temp[7].xyz, src0.xyz, src1.xyz, src0.000
> 23: src0.xyz = temp[5], src1.xyz = const[4], src2.xyz = temp[7]
>     MAD temp[7].xyz, src0.xyz, src1.xyz, src2.xyz
> 24: 
>     MAD color[0].w, src0.0, src0.1, src0.0
> 25: src0.xyz = const[5], src1.xyz = temp[7]
>     MAD temp[10].xyz, src0.xyz, src0.111, -src1.xyz
> 26: src0.xyz = input[2], src1.xyz = temp[10], src2.xyz = temp[7]
>     MAD color[0].xyz, src0.xxx, src1.xyz, src2.xyz
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[11].x, input[4].xy__, 2D[3];
>  2: TEX temp[1], input[6].xy__, 2D[0];
>  3: src0.xyz = temp[1], src1.xyz = input[1]
>     MAD temp[5].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.0, src0.1, src0.0
>  4: src0.xyz = temp[5], src0.w = input[0], src1.xyz = const[3]
>     MAD temp[6].xyz, src0.xyz, src1.xyz, src0.000
>     RCP temp[8].w, src0.w
>  5: src0.xyz = input[0], src0.w = temp[8]
>     MAD temp[8].xy, src0.xy_, src0.ww_, src0.000
>  6: src0.xyz = temp[8], src1.xyz = const[9], src2.xyz = const[10]
>     MAD temp[8].xy, src0.xy_, src1.xy_, src2.xy_
>  7: src0.xyz = const[6], src1.xyz = temp[8]
>     MAD temp[0].y, src0._y_, src0.111, -src1._y_
>  8: src0.xyz = temp[8]
>     MAD temp[0].x, src0.x__, src0.111, src0.000
>  9: src0.xyz = temp[0], src1.xyz = const[7]
>     MAD temp[2].xy, src0.xy_, src1.xx_, src0.000
> 10: BEGIN_TEX;
> 11: TEX temp[2].w, temp[2].xy__, 2D[1];
> 12: src0.w = input[7], src1.w = temp[2]
>     MAD temp[9].w, src0.w, src0.1, -src1.w
> 13: src0.w = temp[9]
>     CMP temp[2].w, src0.0, src0.1, src0.w
> 14: src0.w = temp[2]
>     MAD temp[3].xyz, -src0.www, src0.111, src0.000
>     MAD temp[3].w, -src0.w, src0.1, src0.0
> 15: src0.w = temp[1], src1.w = const[1]
>     MAD temp[2].w, src0.w, src0.1, -src1.w
> 16: src0.w = temp[2]
>     MAD temp[12].xyz, src0.www, src0.111, src0.000
>     MAD temp[12].w, src0.w, src0.1, src0.0
> 17: BEGIN_TEX;
> 18: KIL temp[12];
> 19: KIL temp[3];
> 20: TXP temp[2].xyz, input[5].xy_w, 2D[2];
> 21: src0.xyz = input[3], src1.xyz = temp[2]
>     MAD temp[4].xyz, src0.zzz, src1.xyz, src0.000
> 22: src0.xyz = temp[4]
>     MAX temp[4].xyz, src0.xyz, src0.000
> 23: src0.xyz = temp[4], src1.xyz = temp[11]
>     MAD temp[4].xyz, src0.xyz, src1.xxx, src0.000
> 24: src0.xyz = temp[4], src1.xyz = temp[6]
>     MAD temp[7].xyz, src0.xyz, src1.xyz, src0.000
> 25: src0.xyz = temp[5], src1.xyz = const[4], src2.xyz = temp[7]
>     MAD temp[7].xyz, src0.xyz, src1.xyz, src2.xyz
> 26: src0.xyz = const[5], src1.xyz = temp[7]
>     MAD temp[10].xyz, src0.xyz, src0.111, -src1.xyz
> 27: src0.xyz = input[2], src1.xyz = temp[10], src2.xyz = temp[7]
>     MAD color[0].xyz, src0.xxx, src1.xyz, src2.xyz
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[10].x, temp[3].xy__, 2D[3];
>  2: TEX temp[3], temp[5].xy__, 2D[0];
>  3: src0.xyz = temp[3], src1.xyz = temp[0]
>     MAD temp[8].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.0, src0.1, src0.0
>  4: src0.xyz = temp[8], src0.w = temp[7], src1.xyz = const[3]
>     MAD temp[9].xyz, src0.xyz, src1.xyz, src0.000
>     RCP temp[5].w, src0.w
>  5: src0.xyz = temp[7], src0.w = temp[5]
>     MAD temp[5].xy, src0.xy_, src0.ww_, src0.000
>  6: src0.xyz = temp[5], src1.xyz = const[9], src2.xyz = const[10]
>     MAD temp[5].xy, src0.xy_, src1.xy_, src2.xy_
>  7: src0.xyz = const[6], src1.xyz = temp[5]
>     MAD temp[0].y, src0._y_, src0.111, -src1._y_
>  8: src0.xyz = temp[5]
>     MAD temp[0].x, src0.x__, src0.111, src0.000
>  9: src0.xyz = temp[0], src1.xyz = const[7]
>     MAD temp[0].xy, src0.xy_, src1.xx_, src0.000
> 10: BEGIN_TEX;
> 11: TEX temp[0].w, temp[0].xy__, 2D[1];
> 12: src0.w = temp[6], src1.w = temp[0]
>     MAD temp[5].w, src0.w, src0.1, -src1.w
> 13: src0.w = temp[5]
>     CMP temp[0].w, src0.0, src0.1, src0.w
> 14: src0.w = temp[0]
>     MAD temp[5].xyz, -src0.www, src0.111, src0.000
>     MAD temp[5].w, -src0.w, src0.1, src0.0
> 15: src0.w = temp[3], src1.w = const[1]
>     MAD temp[0].w, src0.w, src0.1, -src1.w
> 16: src0.w = temp[0]
>     MAD temp[3].xyz, src0.www, src0.111, src0.000
>     MAD temp[3].w, src0.w, src0.1, src0.0
> 17: BEGIN_TEX;
> 18: KIL temp[3];
> 19: KIL temp[5];
> 20: TXP temp[0].xyz, temp[4].xy_w, 2D[2];
> 21: src0.xyz = temp[2], src1.xyz = temp[0]
>     MAD temp[0].xyz, src0.zzz, src1.xyz, src0.000
> 22: src0.xyz = temp[0]
>     MAX temp[0].xyz, src0.xyz, src0.000
> 23: src0.xyz = temp[0], src1.xyz = temp[10]
>     MAD temp[0].xyz, src0.xyz, src1.xxx, src0.000
> 24: src0.xyz = temp[0], src1.xyz = temp[9]
>     MAD temp[0].xyz, src0.xyz, src1.xyz, src0.000
> 25: src0.xyz = temp[8], src1.xyz = const[4], src2.xyz = temp[0]
>     MAD temp[0].xyz, src0.xyz, src1.xyz, src2.xyz
> 26: src0.xyz = const[5], src1.xyz = temp[0]
>     MAD temp[2].xyz, src0.xyz, src0.111, -src1.xyz
> 27: src0.xyz = temp[1], src1.xyz = temp[2], src2.xyz = temp[0]
>     MAD color[0].xyz, src0.xxx, src1.xyz, src2.xyz
>pc=26*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 6, tex_end: 1  (code_addr: 00420180)
>  TEX:
>    TEX t10, t3, texture[3]   (00009a83)
>    TEX t3, t5, texture[0]   (000080c5)
>  0: xyz:  t3  t0  t0 -> t8.xyz               (03a00003)
>       w:  t0  t0  t0 -> o0.w                 (01000000)
>     xyz:   t3.xyz   t0.xyz      0.0    op: 00050200
>       w:      0.0      1.0      0.0    op: 00040890
>  1: xyz:  t8  c3  t0 -> t9.xyz               (03a408c8)
>       w:  t7  t0  t0 -> t5.w                 (00940007)
>     xyz:   t8.xyz   c3.xyz      0.0    op: 00050200
>       w:     t7.w     t8.x     t8.x    op: 05000009
>  2: xyz:  t7  t0  t0 -> t5.xy                (01940007)
>       w:  t5  t0  t0 ->                      (00000005)
>     xyz:   t7.xyz   t5.www      0.0    op: 00050600
>       w:     t7.x     t7.x     t7.x    op: 00000000
>  3: xyz:  t5  c9 c10 -> t5.xy                (0196aa45)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t5.xyz   c9.xyz  c10.xyz    op: 00020200
>       w:     t5.x     t5.x     t5.x    op: 00000000
>  4: xyz:  c6  t5  t0 -> t0.y                 (01000166)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   c6.xyz      1.0  -t5.xyz    op: 00090a80
>       w:     c6.x     c6.x     c6.x    op: 00000000
>  5: xyz:  t5  t0  t0 -> t0.x                 (00800005)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t5.xyz      1.0      0.0    op: 00050a80
>       w:     t5.x     t5.x     t5.x    op: 00000000
>  6: xyz:  t0  c7  t0 -> t0.xy                (018009c0)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xyz   c7.xxx      0.0    op: 00050280
>       w:     t0.x     t0.x     t0.x    op: 00000000
>NODE 1: alu_offset: 7, tex_offset: 2, alu_end: 4, tex_end: 0  (code_addr: 00002107)
>  TEX:
>    TEX t0, t0, texture[1]   (00008800)
>  7: xyz:  t0  t0  t0 ->                      (00000000)
>       w:  t6  t0  t0 -> t5.w                 (00940006)
>     xyz:   t0.xxx   t0.xxx   t0.xxx    op: 00004081
>       w:     t6.w      1.0    -t0.w    op: 000a8889
>  8: xyz:  t0  t0  t0 ->                      (00000000)
>       w:  t5  t0  t0 -> t0.w                 (00800005)
>     xyz:   t0.xxx   t0.xxx   t0.xxx    op: 00004081
>       w:      0.0      1.0     t5.w    op: 03024890
>  9: xyz:  t0  t0  t0 -> t5.xyz               (03940000)
>       w:  t0  t0  t0 -> t5.w                 (00940000)
>     xyz:  -t0.www      1.0      0.0    op: 00050aac
>       w:    -t0.w      1.0      0.0    op: 000408a9
> 10: xyz:  t0  t0  t0 ->                      (00000000)
>       w:  t3  c1  t0 -> t0.w                 (00800843)
>     xyz:   t0.xxx   t0.xxx   t0.xxx    op: 00004081
>       w:     t3.w      1.0    -c1.w    op: 000a8889
> 11: xyz:  t0  t0  t0 -> t3.xyz               (038c0000)
>       w:  t0  t0  t0 -> t3.w                 (008c0000)
>     xyz:   t0.www      1.0      0.0    op: 00050a8c
>       w:     t0.w      1.0      0.0    op: 00040889
>NODE 2: alu_offset: 12, tex_offset: 3, alu_end: 6, tex_end: 2  (code_addr: 0044318c)
>  TEX:
>    KIL t0, t3, texture[0]   (00010003)
>    KIL t0, t5, texture[0]   (00010005)
>    TXP t0, t4, texture[2]   (00019004)
> 12: xyz:  t2  t0  t0 -> t0.xyz               (03800002)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.zzz   t0.xyz      0.0    op: 00050203
>       w:     t2.x     t2.x     t2.x    op: 00000000
> 13: xyz:  t0  t0  t0 -> t0.xyz               (03800000)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xyz      0.0   t0.xxx    op: 02804a00
>       w:     t0.x     t0.x     t0.x    op: 00000000
> 14: xyz:  t0 t10  t0 -> t0.xyz               (03800280)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xyz  t10.xxx      0.0    op: 00050280
>       w:     t0.x     t0.x     t0.x    op: 00000000
> 15: xyz:  t0  t9  t0 -> t0.xyz               (03800240)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xyz   t9.xyz      0.0    op: 00050200
>       w:     t0.x     t0.x     t0.x    op: 00000000
> 16: xyz:  t8  c4  t0 -> t0.xyz               (03800908)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t8.xyz   c4.xyz   t0.xyz    op: 00020200
>       w:     t8.x     t8.x     t8.x    op: 00000000
> 17: xyz:  c5  t0  t0 -> t2.xyz               (03880025)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   c5.xyz      1.0  -t0.xyz    op: 00090a80
>       w:     c5.x     c5.x     c5.x    op: 00000000
> 18: xyz:  t1  t2  t0 -> o0.xyz               (1c000081)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xxx   t2.xyz   t0.xyz    op: 00020201
>       w:     t1.x     t1.x     t1.x    op: 00000000
>r300: Initial fragment program
>FRAG
>DCL IN[0], COLOR[1], LINEAR
>DCL IN[1], GENERIC[0], PERSPECTIVE
>DCL IN[2], GENERIC[1], PERSPECTIVE
>DCL IN[3], GENERIC[7], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL SAMP[1]
>DCL SAMP[2]
>DCL CONST[2]
>DCL CONST[5..6]
>DCL TEMP[0..10]
>IMM FLT32 {    0.0000,     0.5000,     1.0000,     2.0000}
>IMM FLT32 {  -32.0000,    16.0000,     0.0000,     2.0000}
>IMM FLT32 {100000.0000,     0.3500,     2.0000,     4.0000}
>IMM FLT32 {    2.0000,     4.0000,     8.0000,    16.0000}
>  0: SUB TEMP[0], IMM[0].yyyy, IN[1]
>  1: SUB TEMP[1], IMM[0].yyyy, IN[2]
>  2: ABS TEMP[0], TEMP[0]
>  3: ABS TEMP[1], TEMP[1]
>  4: MAX TEMP[0].xy, TEMP[0].xzwy, TEMP[0].ywxz
>  5: MAX TEMP[0].zw, TEMP[1].zwxz, TEMP[1].xyyw
>  6: MAD_SAT TEMP[2].yw, TEMP[0].zxwy, IMM[1].xxxx, IMM[1].yyyy
>  7: MUL TEMP[2], TEMP[2], CONST[2]
>  8: SUB TEMP[3], IMM[0].zzzz, TEMP[2]
>  9: MOV TEMP[4], IMM[0].xxxx
> 10: MOV TEMP[5], IN[3]
> 11: MUL_SAT TEMP[6], TEMP[2], IMM[2].xxxx
> 12: MUL TEMP[7], IMM[3], TEMP[6]
> 13: MUL_SAT TEMP[6].y, TEMP[3].yyyy, IMM[2].xxxx
> 14: MUL TEMP[0].xy, TEMP[6].yyyy, TEMP[5]
> 15: MUL TEMP[8].xy, TEMP[5], TEMP[7].yyyy
> 16: MUL TEMP[9].xy, TEMP[5], TEMP[7].wwww
> 17: TEX TEMP[0], TEMP[0], SAMP[0], 2D
> 18: TEX TEMP[8], TEMP[8], SAMP[1], 2D
> 19: TEX TEMP[9], TEMP[9], SAMP[2], 2D
> 20: MOV TEMP[4], TEMP[0]
> 21: MUL TEMP[4], TEMP[4], TEMP[3].yyyy
> 22: MAD TEMP[4], TEMP[8], TEMP[2].yyyy, TEMP[4]
> 23: MUL TEMP[4], TEMP[4], TEMP[3].wwww
> 24: MAD TEMP[4], TEMP[9], TEMP[2].wwww, TEMP[4]
> 25: LRP TEMP[10], IMM[2].yyyy, CONST[5], IMM[0].zzzz
> 26: MUL TEMP[4].xyz, TEMP[4], TEMP[10]
> 27: LRP OUT[0].xyz, IN[0].xxxx, CONST[6], TEMP[4]
> 28: MOV OUT[0].w, IMM[0].xxxx
> 29: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: SUB temp[0], const[7].yyyy, input[1];
>  1: SUB temp[1], const[7].yyyy, input[2];
>  2: ABS temp[0], temp[0];
>  3: ABS temp[1], temp[1];
>  4: MAX temp[0].xy, temp[0].xzwy, temp[0].ywxz;
>  5: MAX temp[0].zw, temp[1].zwxz, temp[1].xyyw;
>  6: MAD_SAT temp[2].yw, temp[0].zxwy, const[8].xxxx, const[8].yyyy;
>  7: MUL temp[2], temp[2], const[2];
>  8: SUB temp[3], const[7].zzzz, temp[2];
>  9: MOV temp[4], const[7].xxxx;
> 10: MOV temp[5], input[3];
> 11: MUL_SAT temp[6], temp[2], const[9].xxxx;
> 12: MUL temp[7], const[10], temp[6];
> 13: MUL_SAT temp[6].y, temp[3].yyyy, const[9].xxxx;
> 14: MUL temp[0].xy, temp[6].yyyy, temp[5];
> 15: MUL temp[8].xy, temp[5], temp[7].yyyy;
> 16: MUL temp[9].xy, temp[5], temp[7].wwww;
> 17: TEX temp[0], temp[0], 2D[0];
> 18: TEX temp[8], temp[8], 2D[1];
> 19: TEX temp[9], temp[9], 2D[2];
> 20: MOV temp[4], temp[0];
> 21: MUL temp[4], temp[4], temp[3].yyyy;
> 22: MAD temp[4], temp[8], temp[2].yyyy, temp[4];
> 23: MUL temp[4], temp[4], temp[3].wwww;
> 24: MAD temp[4], temp[9], temp[2].wwww, temp[4];
> 25: LRP temp[10], const[9].yyyy, const[5], const[7].zzzz;
> 26: MUL temp[4].xyz, temp[4], temp[10];
> 27: LRP output[0].xyz, input[0].xxxx, const[6], temp[4];
> 28: MOV output[0].w, const[7].xxxx;
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: SUB temp[0], const[7].yyyy, input[1];
>  1: SUB temp[1], const[7].yyyy, input[2];
>  2: ABS temp[0], temp[0];
>  3: ABS temp[1], temp[1];
>  4: MAX temp[0].xy, temp[0].xzwy, temp[0].ywxz;
>  5: MAX temp[0].zw, temp[1].zwxz, temp[1].xyyw;
>  6: MAD_SAT temp[2].yw, temp[0].zxwy, const[8].xxxx, const[8].yyyy;
>  7: MUL temp[2], temp[2], const[2];
>  8: SUB temp[3], const[7].zzzz, temp[2];
>  9: MOV temp[4], const[7].xxxx;
> 10: MOV temp[5], input[3];
> 11: MUL_SAT temp[6], temp[2], const[9].xxxx;
> 12: MUL temp[7], const[10], temp[6];
> 13: MUL_SAT temp[6].y, temp[3].yyyy, const[9].xxxx;
> 14: MUL temp[0].xy, temp[6].yyyy, temp[5];
> 15: MUL temp[8].xy, temp[5], temp[7].yyyy;
> 16: MUL temp[9].xy, temp[5], temp[7].wwww;
> 17: TEX temp[0], temp[0], 2D[0];
> 18: TEX temp[8], temp[8], 2D[1];
> 19: TEX temp[9], temp[9], 2D[2];
> 20: MOV temp[4], temp[0];
> 21: MUL temp[4], temp[4], temp[3].yyyy;
> 22: MAD temp[4], temp[8], temp[2].yyyy, temp[4];
> 23: MUL temp[4], temp[4], temp[3].wwww;
> 24: MAD temp[4], temp[9], temp[2].wwww, temp[4];
> 25: LRP temp[10], const[9].yyyy, const[5], const[7].zzzz;
> 26: MUL temp[4].xyz, temp[4], temp[10];
> 27: LRP output[0].xyz, input[0].xxxx, const[6], temp[4];
> 28: MOV output[0].w, const[7].xxxx;
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: SUB temp[0], const[7].yyyy, input[1];
>  1: SUB temp[1], const[7].yyyy, input[2];
>  2: ABS temp[0], temp[0];
>  3: ABS temp[1], temp[1];
>  4: MAX temp[0].xy, temp[0].xzwy, temp[0].ywxz;
>  5: MAX temp[0].zw, temp[1].zwxz, temp[1].xyyw;
>  6: MAD_SAT temp[2].yw, temp[0].zxwy, const[8].xxxx, const[8].yyyy;
>  7: MUL temp[2], temp[2], const[2];
>  8: SUB temp[3], const[7].zzzz, temp[2];
>  9: MOV temp[4], const[7].xxxx;
> 10: MOV temp[5], input[3];
> 11: MUL_SAT temp[6], temp[2], const[9].xxxx;
> 12: MUL temp[7], const[10], temp[6];
> 13: MUL_SAT temp[6].y, temp[3].yyyy, const[9].xxxx;
> 14: MUL temp[0].xy, temp[6].yyyy, temp[5];
> 15: MUL temp[8].xy, temp[5], temp[7].yyyy;
> 16: MUL temp[9].xy, temp[5], temp[7].wwww;
> 17: TEX temp[0], temp[0], 2D[0];
> 18: TEX temp[8], temp[8], 2D[1];
> 19: TEX temp[9], temp[9], 2D[2];
> 20: MOV temp[4], temp[0];
> 21: MUL temp[4], temp[4], temp[3].yyyy;
> 22: MAD temp[4], temp[8], temp[2].yyyy, temp[4];
> 23: MUL temp[4], temp[4], temp[3].wwww;
> 24: MAD temp[4], temp[9], temp[2].wwww, temp[4];
> 25: LRP temp[10], const[9].yyyy, const[5], const[7].zzzz;
> 26: MUL temp[4].xyz, temp[4], temp[10];
> 27: LRP output[0].xyz, input[0].xxxx, const[6], temp[4];
> 28: MOV output[0].w, const[7].xxxx;
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: ADD temp[0], const[7].yyyy, -input[1];
>  1: ADD temp[1], const[7].yyyy, -input[2];
>  2: MOV temp[0], |temp[0]|;
>  3: MOV temp[1], |temp[1]|;
>  4: MAX temp[0].xy, temp[0].xzwy, temp[0].ywxz;
>  5: MAX temp[0].zw, temp[1].zwxz, temp[1].xyyw;
>  6: MAD_SAT temp[2].yw, temp[0].zxwy, const[8].xxxx, const[8].yyyy;
>  7: MUL temp[2], temp[2], const[2];
>  8: ADD temp[3], const[7].zzzz, -temp[2];
>  9: MOV temp[4], const[7].xxxx;
> 10: MOV temp[5], input[3];
> 11: MUL_SAT temp[6], temp[2], const[9].xxxx;
> 12: MUL temp[7], const[10], temp[6];
> 13: MUL_SAT temp[6].y, temp[3].yyyy, const[9].xxxx;
> 14: MUL temp[0].xy, temp[6].yyyy, temp[5];
> 15: MUL temp[8].xy, temp[5], temp[7].yyyy;
> 16: MUL temp[9].xy, temp[5], temp[7].wwww;
> 17: TEX temp[0], temp[0], 2D[0];
> 18: TEX temp[8], temp[8], 2D[1];
> 19: TEX temp[9], temp[9], 2D[2];
> 20: MOV temp[4], temp[0];
> 21: MUL temp[4], temp[4], temp[3].yyyy;
> 22: MAD temp[4], temp[8], temp[2].yyyy, temp[4];
> 23: MUL temp[4], temp[4], temp[3].wwww;
> 24: MAD temp[4], temp[9], temp[2].wwww, temp[4];
> 25: ADD temp[11], const[5], -const[7].zzzz;
> 26: MAD temp[10], const[9].yyyy, temp[11], const[7].zzzz;
> 27: MUL temp[4].xyz, temp[4], temp[10];
> 28: ADD temp[12], const[6], -temp[4];
> 29: MAD output[0].xyz, input[0].xxxx, temp[12], temp[4];
> 30: MOV output[0].w, const[7].xxxx;
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: ADD temp[0], const[7].yyyy, -input[1];
>  1: ADD temp[1], const[7].yyyy, -input[2];
>  2: MOV temp[0], |temp[0]|;
>  3: MOV temp[1], |temp[1]|;
>  4: MAX temp[0].xy, temp[0].xzwy, temp[0].ywxz;
>  5: MAX temp[0].zw, temp[1].zwxz, temp[1].xyyw;
>  6: MAD_SAT temp[2].yw, temp[0].zxwy, const[8].xxxx, const[8].yyyy;
>  7: MUL temp[2], temp[2], const[2];
>  8: ADD temp[3], const[7].zzzz, -temp[2];
>  9: MOV temp[4], const[7].xxxx;
> 10: MOV temp[5], input[3];
> 11: MUL_SAT temp[6], temp[2], const[9].xxxx;
> 12: MUL temp[7], const[10], temp[6];
> 13: MUL_SAT temp[6].y, temp[3].yyyy, const[9].xxxx;
> 14: MUL temp[0].xy, temp[6].yyyy, temp[5];
> 15: MUL temp[8].xy, temp[5], temp[7].yyyy;
> 16: MUL temp[9].xy, temp[5], temp[7].wwww;
> 17: TEX temp[0], temp[0], 2D[0];
> 18: TEX temp[8], temp[8], 2D[1];
> 19: TEX temp[9], temp[9], 2D[2];
> 20: MOV temp[4], temp[0];
> 21: MUL temp[4], temp[4], temp[3].yyyy;
> 22: MAD temp[4], temp[8], temp[2].yyyy, temp[4];
> 23: MUL temp[4], temp[4], temp[3].wwww;
> 24: MAD temp[4], temp[9], temp[2].wwww, temp[4];
> 25: ADD temp[11], const[5], -const[7].zzzz;
> 26: MAD temp[10], const[9].yyyy, temp[11], const[7].zzzz;
> 27: MUL temp[4].xyz, temp[4], temp[10];
> 28: ADD temp[12], const[6], -temp[4];
> 29: MAD output[0].xyz, input[0].xxxx, temp[12], temp[4];
> 30: MOV output[0].w, const[7].xxxx;
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: ADD temp[0], const[7].yyyy, -input[1];
>  1: MOV temp[0], |temp[0]|;
>  2: MAX temp[0].xy, temp[0].xz__, temp[0].yw__;
>  3: MAD_SAT temp[2].yw, temp[0]._x_y, const[8]._x_x, const[8]._y_y;
>  4: MUL temp[2].yw, temp[2]._y_w, const[2]._y_w;
>  5: ADD temp[3].yw, const[7]._z_z, -temp[2]._y_w;
>  6: MOV temp[5].xy, input[3].xy__;
>  7: MUL_SAT temp[6].yw, temp[2]._y_w, const[9]._x_x;
>  8: MUL temp[7].yw, const[10]._y_w, temp[6]._y_w;
>  9: MUL_SAT temp[6].y, temp[3]._y__, const[9]._x__;
> 10: MUL temp[0].xy, temp[6].yy__, temp[5].xy__;
> 11: MUL temp[8].xy, temp[5].xy__, temp[7].yy__;
> 12: MUL temp[9].xy, temp[5].xy__, temp[7].ww__;
> 13: TEX temp[0].xyz, temp[0].xy__, 2D[0];
> 14: TEX temp[8].xyz, temp[8].xy__, 2D[1];
> 15: TEX temp[9].xyz, temp[9].xy__, 2D[2];
> 16: MOV temp[4].xyz, temp[0].xyz_;
> 17: MUL temp[4].xyz, temp[4].xyz_, temp[3].yyy_;
> 18: MAD temp[4].xyz, temp[8].xyz_, temp[2].yyy_, temp[4].xyz_;
> 19: MUL temp[4].xyz, temp[4].xyz_, temp[3].www_;
> 20: MAD temp[4].xyz, temp[9].xyz_, temp[2].www_, temp[4].xyz_;
> 21: ADD temp[11].xyz, const[5].xyz_, -const[7].zzz_;
> 22: MAD temp[10].xyz, const[9].yyy_, temp[11].xyz_, const[7].zzz_;
> 23: MUL temp[4].xyz, temp[4].xyz_, temp[10].xyz_;
> 24: ADD temp[12].xyz, const[6].xyz_, -temp[4].xyz_;
> 25: MAD output[0].xyz, input[0].xxx_, temp[12].xyz_, temp[4].xyz_;
> 26: MOV output[0].w, const[7].___x;
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: ADD temp[0], none.HHHH, -input[1];
>  1: MAX temp[0].xy, |temp[0].xz__|, |temp[0].yw__|;
>  2: MAD_SAT temp[2].yw, temp[0]._x_y, const[8]._x_x, const[8]._y_y;
>  3: MUL temp[2].yw, temp[2]._y_w, const[2]._y_w;
>  4: ADD temp[3].yw, none._1_1, -temp[2]._y_w;
>  5: MUL_SAT temp[6].yw, temp[2]._y_w, const[9]._x_x;
>  6: MUL temp[7].yw, const[10]._y_w, temp[6]._y_w;
>  7: MUL_SAT temp[6].y, temp[3]._y__, const[9]._x__;
>  8: MUL temp[0].xy, temp[6].yy__, input[3].xy__;
>  9: MUL temp[8].xy, input[3].xy__, temp[7].yy__;
> 10: MUL temp[9].xy, input[3].xy__, temp[7].ww__;
> 11: TEX temp[0].xyz, temp[0].xy__, 2D[0];
> 12: TEX temp[8].xyz, temp[8].xy__, 2D[1];
> 13: TEX temp[9].xyz, temp[9].xy__, 2D[2];
> 14: MUL temp[4].xyz, temp[0].xyz_, temp[3].yyy_;
> 15: MAD temp[4].xyz, temp[8].xyz_, temp[2].yyy_, temp[4].xyz_;
> 16: MUL temp[4].xyz, temp[4].xyz_, temp[3].www_;
> 17: MAD temp[4].xyz, temp[9].xyz_, temp[2].www_, temp[4].xyz_;
> 18: ADD temp[11].xyz, const[5].xyz_, -none.111_;
> 19: MAD temp[10].xyz, const[9].yyy_, temp[11].xyz_, none.111_;
> 20: MUL temp[4].xyz, temp[4].xyz_, temp[10].xyz_;
> 21: ADD temp[12].xyz, const[6].xyz_, -temp[4].xyz_;
> 22: MAD output[0].xyz, input[0].xxx_, temp[12].xyz_, temp[4].xyz_;
> 23: MOV output[0].w, none.___0;
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: ADD temp[0], none.HHHH, -input[1];
>  1: MOV temp[1].x, |temp[0].x___|;
>  2: MOV temp[1].y, |temp[0]._z__|;
>  3: MOV temp[5].x, |temp[0].y___|;
>  4: MOV temp[5].y, |temp[0]._w__|;
>  5: MAX temp[0].xy, temp[1].xy__, temp[5].xy__;
>  6: MAD_SAT temp[2].yw, temp[0]._x_y, const[8]._x_x, const[8]._y_y;
>  7: MUL temp[2].yw, temp[2]._y_w, const[2]._y_w;
>  8: ADD temp[3].yw, none._1_1, -temp[2]._y_w;
>  9: MUL_SAT temp[6].yw, temp[2]._y_w, const[9]._x_x;
> 10: MUL temp[7].yw, const[10]._y_w, temp[6]._y_w;
> 11: MUL_SAT temp[6].y, temp[3]._y__, const[9]._x__;
> 12: MUL temp[0].xy, temp[6].yy__, input[3].xy__;
> 13: MUL temp[8].xy, input[3].xy__, temp[7].yy__;
> 14: MUL temp[9].xy, input[3].xy__, temp[7].ww__;
> 15: TEX temp[0].xyz, temp[0].xy__, 2D[0];
> 16: TEX temp[8].xyz, temp[8].xy__, 2D[1];
> 17: TEX temp[9].xyz, temp[9].xy__, 2D[2];
> 18: MUL temp[4].xyz, temp[0].xyz_, temp[3].yyy_;
> 19: MAD temp[4].xyz, temp[8].xyz_, temp[2].yyy_, temp[4].xyz_;
> 20: MUL temp[4].xyz, temp[4].xyz_, temp[3].www_;
> 21: MAD temp[4].xyz, temp[9].xyz_, temp[2].www_, temp[4].xyz_;
> 22: ADD temp[11].xyz, const[5].xyz_, -none.111_;
> 23: MAD temp[10].xyz, const[9].yyy_, temp[11].xyz_, none.111_;
> 24: MUL temp[4].xyz, temp[4].xyz_, temp[10].xyz_;
> 25: ADD temp[12].xyz, const[6].xyz_, -temp[4].xyz_;
> 26: MAD output[0].xyz, input[0].xxx_, temp[12].xyz_, temp[4].xyz_;
> 27: MOV output[0].w, none.___0;
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: src0.xyz = input[1], src0.w = input[1]
>     MAD temp[0].xyz, src0.HHH, src0.111, -src0.xyz
>     MAD temp[0].w, src0.H, src0.1, -src0.w
>  1: src0.xyz = temp[0]
>     MAD temp[1].x, |src0.x__|, src0.111, src0.000
>  2: src0.xyz = temp[0]
>     MAD temp[1].y, |src0._z_|, src0.111, src0.000
>  3: src0.xyz = temp[0]
>     MAD temp[5].x, |src0.y__|, src0.111, src0.000
>  4: src0.w = temp[0]
>     MAD temp[5].y, |src0._w_|, src0.111, src0.000
>  5: src0.xyz = temp[1], src1.xyz = temp[5]
>     MAX temp[0].xy, src0.xy_, src1.xy_
>  6: src0.xyz = temp[0], src1.xyz = const[8]
>     MAD_SAT temp[2].y, src0._x_, src1._x_, src1._y_
>     MAD_SAT temp[2].w, src0.y, src1.x, src1.y
>  7: src0.xyz = temp[2], src0.w = temp[2], src1.xyz = const[2], src1.w = const[2]
>     MAD temp[2].y, src0._y_, src1._y_, src0.000
>     MAD temp[2].w, src0.w, src1.w, src0.0
>  8: src0.xyz = temp[2], src0.w = temp[2]
>     MAD temp[3].y, src0._1_, src0.111, -src0._y_
>     MAD temp[3].w, src0.1, src0.1, -src0.w
>  9: src0.xyz = temp[2], src0.w = temp[2], src1.xyz = const[9]
>     MAD_SAT temp[6].y, src0._y_, src1._x_, src0.000
>     MAD_SAT temp[6].w, src0.w, src1.x, src0.0
> 10: src0.xyz = const[10], src0.w = const[10], src1.xyz = temp[6], src1.w = temp[6]
>     MAD temp[7].y, src0._y_, src1._y_, src0.000
>     MAD temp[7].w, src0.w, src1.w, src0.0
> 11: src0.xyz = temp[3], src1.xyz = const[9]
>     MAD_SAT temp[6].y, src0._y_, src1._x_, src0.000
> 12: src0.xyz = temp[6], src1.xyz = input[3]
>     MAD temp[0].xy, src0.yy_, src1.xy_, src0.000
> 13: src0.xyz = input[3], src1.xyz = temp[7]
>     MAD temp[8].xy, src0.xy_, src1.yy_, src0.000
> 14: src0.xyz = input[3], src0.w = temp[7]
>     MAD temp[9].xy, src0.xy_, src0.ww_, src0.000
> 15: TEX temp[0].xyz, temp[0].xy__, 2D[0];
> 16: TEX temp[8].xyz, temp[8].xy__, 2D[1];
> 17: TEX temp[9].xyz, temp[9].xy__, 2D[2];
> 18: src0.xyz = temp[0], src1.xyz = temp[3]
>     MAD temp[4].xyz, src0.xyz, src1.yyy, src0.000
> 19: src0.xyz = temp[8], src1.xyz = temp[2], src2.xyz = temp[4]
>     MAD temp[4].xyz, src0.xyz, src1.yyy, src2.xyz
> 20: src0.xyz = temp[4], src0.w = temp[3]
>     MAD temp[4].xyz, src0.xyz, src0.www, src0.000
> 21: src0.xyz = temp[9], src0.w = temp[2], src1.xyz = temp[4]
>     MAD temp[4].xyz, src0.xyz, src0.www, src1.xyz
> 22: src0.xyz = const[5]
>     MAD temp[11].xyz, src0.xyz, src0.111, -src0.111
> 23: src0.xyz = const[9], src1.xyz = temp[11]
>     MAD temp[10].xyz, src0.yyy, src1.xyz, src0.111
> 24: src0.xyz = temp[4], src1.xyz = temp[10]
>     MAD temp[4].xyz, src0.xyz, src1.xyz, src0.000
> 25: src0.xyz = const[6], src1.xyz = temp[4]
>     MAD temp[12].xyz, src0.xyz, src0.111, -src1.xyz
> 26: src0.xyz = input[0], src1.xyz = temp[12], src2.xyz = temp[4]
>     MAD color[0].xyz, src0.xxx, src1.xyz, src2.xyz
> 27: 
>     MAD color[0].w, src0.0, src0.1, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: src0.xyz = input[1], src0.w = input[1]
>     MAD temp[0].xyz, src0.HHH, src0.111, -src0.xyz
>     MAD temp[0].w, src0.H, src0.1, -src0.w
>  1: src0.w = temp[0]
>     MAD temp[5].y, |src0._w_|, src0.111, src0.000
>     MAD color[0].w, src0.0, src0.1, src0.0
>  2: src0.xyz = temp[0]
>     MAD temp[1].y, |src0._z_|, src0.111, src0.000
>  3: src0.xyz = temp[0]
>     MAD temp[5].x, |src0.y__|, src0.111, src0.000
>  4: src0.xyz = temp[0]
>     MAD temp[1].x, |src0.x__|, src0.111, src0.000
>  5: src0.xyz = temp[1], src1.xyz = temp[5]
>     MAX temp[0].xy, src0.xy_, src1.xy_
>  6: src0.xyz = temp[0], src1.xyz = const[8]
>     MAD_SAT temp[2].y, src0._x_, src1._x_, src1._y_
>     MAD_SAT temp[2].w, src0.y, src1.x, src1.y
>  7: src0.xyz = temp[2], src0.w = temp[2], src1.xyz = const[2], src1.w = const[2]
>     MAD temp[2].y, src0._y_, src1._y_, src0.000
>     MAD temp[2].w, src0.w, src1.w, src0.0
>  8: src0.xyz = temp[2], src0.w = temp[2]
>     MAD temp[3].y, src0._1_, src0.111, -src0._y_
>     MAD temp[3].w, src0.1, src0.1, -src0.w
>  9: src0.xyz = temp[2], src0.w = temp[2], src1.xyz = const[9]
>     MAD_SAT temp[6].y, src0._y_, src1._x_, src0.000
>     MAD_SAT temp[6].w, src0.w, src1.x, src0.0
> 10: src0.xyz = const[10], src0.w = const[10], src1.xyz = temp[6], src1.w = temp[6]
>     MAD temp[7].y, src0._y_, src1._y_, src0.000
>     MAD temp[7].w, src0.w, src1.w, src0.0
> 11: src0.xyz = input[3], src0.w = temp[7]
>     MAD temp[9].xy, src0.xy_, src0.ww_, src0.000
> 12: src0.xyz = input[3], src1.xyz = temp[7]
>     MAD temp[8].xy, src0.xy_, src1.yy_, src0.000
> 13: src0.xyz = temp[3], src1.xyz = const[9]
>     MAD_SAT temp[6].y, src0._y_, src1._x_, src0.000
> 14: src0.xyz = temp[6], src1.xyz = input[3]
>     MAD temp[0].xy, src0.yy_, src1.xy_, src0.000
> 15: src0.xyz = const[5]
>     MAD temp[11].xyz, src0.xyz, src0.111, -src0.111
> 16: src0.xyz = const[9], src1.xyz = temp[11]
>     MAD temp[10].xyz, src0.yyy, src1.xyz, src0.111
> 17: BEGIN_TEX;
> 18: TEX temp[0].xyz, temp[0].xy__, 2D[0];
> 19: TEX temp[8].xyz, temp[8].xy__, 2D[1];
> 20: TEX temp[9].xyz, temp[9].xy__, 2D[2];
> 21: src0.xyz = temp[0], src1.xyz = temp[3]
>     MAD temp[4].xyz, src0.xyz, src1.yyy, src0.000
> 22: src0.xyz = temp[8], src1.xyz = temp[2], src2.xyz = temp[4]
>     MAD temp[4].xyz, src0.xyz, src1.yyy, src2.xyz
> 23: src0.xyz = temp[4], src0.w = temp[3]
>     MAD temp[4].xyz, src0.xyz, src0.www, src0.000
> 24: src0.xyz = temp[9], src0.w = temp[2], src1.xyz = temp[4]
>     MAD temp[4].xyz, src0.xyz, src0.www, src1.xyz
> 25: src0.xyz = temp[4], src1.xyz = temp[10]
>     MAD temp[4].xyz, src0.xyz, src1.xyz, src0.000
> 26: src0.xyz = const[6], src1.xyz = temp[4]
>     MAD temp[12].xyz, src0.xyz, src0.111, -src1.xyz
> 27: src0.xyz = input[0], src1.xyz = temp[12], src2.xyz = temp[4]
>     MAD color[0].xyz, src0.xxx, src1.xyz, src2.xyz
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: src0.xyz = temp[1], src0.w = temp[1]
>     MAD temp[1].xyz, src0.HHH, src0.111, -src0.xyz
>     MAD temp[1].w, src0.H, src0.1, -src0.w
>  1: src0.w = temp[1]
>     MAD temp[4].y, |src0._w_|, src0.111, src0.000
>     MAD color[0].w, src0.0, src0.1, src0.0
>  2: src0.xyz = temp[1]
>     MAD temp[2].y, |src0._z_|, src0.111, src0.000
>  3: src0.xyz = temp[1]
>     MAD temp[4].x, |src0.y__|, src0.111, src0.000
>  4: src0.xyz = temp[1]
>     MAD temp[2].x, |src0.x__|, src0.111, src0.000
>  5: src0.xyz = temp[2], src1.xyz = temp[4]
>     MAX temp[1].xy, src0.xy_, src1.xy_
>  6: src0.xyz = temp[1], src1.xyz = const[8]
>     MAD_SAT temp[2].y, src0._x_, src1._x_, src1._y_
>     MAD_SAT temp[2].w, src0.y, src1.x, src1.y
>  7: src0.xyz = temp[2], src0.w = temp[2], src1.xyz = const[2], src1.w = const[2]
>     MAD temp[2].y, src0._y_, src1._y_, src0.000
>     MAD temp[2].w, src0.w, src1.w, src0.0
>  8: src0.xyz = temp[2], src0.w = temp[2]
>     MAD temp[4].y, src0._1_, src0.111, -src0._y_
>     MAD temp[4].w, src0.1, src0.1, -src0.w
>  9: src0.xyz = temp[2], src0.w = temp[2], src1.xyz = const[9]
>     MAD_SAT temp[5].y, src0._y_, src1._x_, src0.000
>     MAD_SAT temp[5].w, src0.w, src1.x, src0.0
> 10: src0.xyz = const[10], src0.w = const[10], src1.xyz = temp[5], src1.w = temp[5]
>     MAD temp[6].y, src0._y_, src1._y_, src0.000
>     MAD temp[6].w, src0.w, src1.w, src0.0
> 11: src0.xyz = temp[3], src0.w = temp[6]
>     MAD temp[7].xy, src0.xy_, src0.ww_, src0.000
> 12: src0.xyz = temp[3], src1.xyz = temp[6]
>     MAD temp[6].xy, src0.xy_, src1.yy_, src0.000
> 13: src0.xyz = temp[4], src1.xyz = const[9]
>     MAD_SAT temp[5].y, src0._y_, src1._x_, src0.000
> 14: src0.xyz = temp[5], src1.xyz = temp[3]
>     MAD temp[1].xy, src0.yy_, src1.xy_, src0.000
> 15: src0.xyz = const[5]
>     MAD temp[3].xyz, src0.xyz, src0.111, -src0.111
> 16: src0.xyz = const[9], src1.xyz = temp[3]
>     MAD temp[3].xyz, src0.yyy, src1.xyz, src0.111
> 17: BEGIN_TEX;
> 18: TEX temp[1].xyz, temp[1].xy__, 2D[0];
> 19: TEX temp[6].xyz, temp[6].xy__, 2D[1];
> 20: TEX temp[7].xyz, temp[7].xy__, 2D[2];
> 21: src0.xyz = temp[1], src1.xyz = temp[4]
>     MAD temp[1].xyz, src0.xyz, src1.yyy, src0.000
> 22: src0.xyz = temp[6], src1.xyz = temp[2], src2.xyz = temp[1]
>     MAD temp[1].xyz, src0.xyz, src1.yyy, src2.xyz
> 23: src0.xyz = temp[1], src0.w = temp[4]
>     MAD temp[1].xyz, src0.xyz, src0.www, src0.000
> 24: src0.xyz = temp[7], src0.w = temp[2], src1.xyz = temp[1]
>     MAD temp[1].xyz, src0.xyz, src0.www, src1.xyz
> 25: src0.xyz = temp[1], src1.xyz = temp[3]
>     MAD temp[1].xyz, src0.xyz, src1.xyz, src0.000
> 26: src0.xyz = const[6], src1.xyz = temp[1]
>     MAD temp[2].xyz, src0.xyz, src0.111, -src1.xyz
> 27: src0.xyz = temp[0], src1.xyz = temp[2], src2.xyz = temp[1]
>     MAD color[0].xyz, src0.xxx, src1.xyz, src2.xyz
>pc=27*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 16, tex_end: 0  (code_addr: 00400400)
>  0: xyz:  t1  t0  t0 -> t1.xyz               (03840001)
>       w:  t1  t0  t0 -> t1.w                 (00840001)
>     xyz:      0.5      1.0  -t1.xyz    op: 00080a96
>       w:      0.5      1.0    -t1.w    op: 000a4892
>  1: xyz:  t0  t0  t0 -> t4.y                 (01100000)
>       w:  t1  t0  t0 -> o0.w                 (01000001)
>     xyz: |t1.www|      1.0      0.0    op: 00050acc
>       w:      0.0      1.0      0.0    op: 00040890
>  2: xyz:  t1  t0  t0 -> t2.y                 (01080001)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz: |t1.zzz|      1.0      0.0    op: 00050ac3
>       w:     t1.x     t1.x     t1.x    op: 00000000
>  3: xyz:  t1  t0  t0 -> t4.x                 (00900001)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz: |t1.yyy|      1.0      0.0    op: 00050ac2
>       w:     t1.x     t1.x     t1.x    op: 00000000
>  4: xyz:  t1  t0  t0 -> t2.x                 (00880001)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz: |t1.xyz|      1.0      0.0    op: 00050ac0
>       w:     t1.x     t1.x     t1.x    op: 00000000
>  5: xyz:  t2  t4  t0 -> t1.xy                (01840102)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz   t4.xyz   t2.xxx    op: 02804200
>       w:     t2.x     t2.x     t2.x    op: 00000000
>  6: xyz:  t1  c8  t0 -> t2.y                 (01080a01)
>       w:  t0  t0  t0 -> t2.w                 (00880000)
>     xyz:   t1.xxx   c8.xxx   c8.xyz    op: 40010281
>       w:     t1.y     c8.x     c8.y    op: 40010181
>  7: xyz:  t2  c2  t0 -> t2.y                 (01080882)
>       w:  t2  c2  t0 -> t2.w                 (00880882)
>     xyz:   t2.xyz   c2.xyz      0.0    op: 00050200
>       w:     t2.w     c2.w      0.0    op: 00040509
>  8: xyz:  t2  t0  t0 -> t4.y                 (01100002)
>       w:  t2  t0  t0 -> t4.w                 (00900002)
>     xyz:      1.0      1.0  -t2.xyz    op: 00080a95
>       w:      1.0      1.0    -t2.w    op: 000a4891
>  9: xyz:  t2  c9  t0 -> t5.y                 (01140a42)
>       w:  t2  t0  t0 -> t5.w                 (00940002)
>     xyz:   t2.xyz   c9.xxx      0.0    op: 40050280
>       w:     t2.w     c9.x      0.0    op: 40040189
> 10: xyz: c10  t5  t0 -> t6.y                 (0118016a)
>       w: c10  t5  t0 -> t6.w                 (0098016a)
>     xyz:  c10.xyz   t5.xyz      0.0    op: 00050200
>       w:    c10.w     t5.w      0.0    op: 00040509
> 11: xyz:  t3  t0  t0 -> t7.xy                (019c0003)
>       w:  t6  t0  t0 ->                      (00000006)
>     xyz:   t3.xyz   t6.www      0.0    op: 00050600
>       w:     t3.x     t3.x     t3.x    op: 00000000
> 12: xyz:  t3  t6  t0 -> t6.xy                (01980183)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t3.xyz   t6.yyy      0.0    op: 00050300
>       w:     t3.x     t3.x     t3.x    op: 00000000
> 13: xyz:  t4  c9  t0 -> t5.y                 (01140a44)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t4.xyz   c9.xxx      0.0    op: 40050280
>       w:     t4.x     t4.x     t4.x    op: 00000000
> 14: xyz:  t5  t3  t0 -> t1.xy                (018400c5)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t5.yyy   t3.xyz      0.0    op: 00050202
>       w:     t5.x     t5.x     t5.x    op: 00000000
> 15: xyz:  c5  t0  t0 -> t3.xyz               (038c0025)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   c5.xyz      1.0     -1.0    op: 000d4a80
>       w:     c5.x     c5.x     c5.x    op: 00000000
> 16: xyz:  c9  t3  t0 -> t3.xyz               (038c00e9)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   c9.yyy   t3.xyz      1.0    op: 00054202
>       w:     c9.x     c9.x     c9.x    op: 00000000
>NODE 1: alu_offset: 17, tex_offset: 0, alu_end: 6, tex_end: 2  (code_addr: 00440191)
>  TEX:
>    TEX t1, t1, texture[0]   (00008041)
>    TEX t6, t6, texture[1]   (00008986)
>    TEX t7, t7, texture[2]   (000091c7)
> 17: xyz:  t1  t4  t0 -> t1.xyz               (03840101)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   t4.yyy      0.0    op: 00050300
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 18: xyz:  t6  t2  t1 -> t1.xyz               (03841086)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t6.xyz   t2.yyy   t1.xyz    op: 00020300
>       w:     t6.x     t6.x     t6.x    op: 00000000
> 19: xyz:  t1  t0  t0 -> t1.xyz               (03840001)
>       w:  t4  t0  t0 ->                      (00000004)
>     xyz:   t1.xyz   t4.www      0.0    op: 00050600
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 20: xyz:  t7  t1  t0 -> t1.xyz               (03840047)
>       w:  t2  t0  t0 ->                      (00000002)
>     xyz:   t7.xyz   t2.www   t1.xyz    op: 00010600
>       w:     t7.x     t7.x     t7.x    op: 00000000
> 21: xyz:  t1  t3  t0 -> t1.xyz               (038400c1)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   t3.xyz      0.0    op: 00050200
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 22: xyz:  c6  t1  t0 -> t2.xyz               (03880066)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   c6.xyz      1.0  -t1.xyz    op: 00090a80
>       w:     c6.x     c6.x     c6.x    op: 00000000
> 23: xyz:  t0  t2  t1 -> o0.xyz               (1c001080)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xxx   t2.xyz   t1.xyz    op: 00020201
>       w:     t0.x     t0.x     t0.x    op: 00000000
>r300: Initial fragment program
>FRAG
>DCL IN[0], COLOR, LINEAR
>DCL IN[1], COLOR[1], LINEAR
>DCL IN[2], GENERIC[0], PERSPECTIVE
>DCL IN[3], GENERIC[2], PERSPECTIVE
>DCL IN[4], GENERIC[3], PERSPECTIVE
>DCL IN[5], GENERIC[4], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL SAMP[1]
>DCL SAMP[2]
>DCL CONST[1..3]
>DCL TEMP[0..6]
>IMM FLT32 {    0.3000,     0.5900,     0.1100,     0.0000}
>  0: TEX TEMP[0], IN[5], SAMP[0], 2D
>  1: MOV TEMP[1], TEMP[2]
>  2: TXP TEMP[2], IN[4], SAMP[1], 2D
>  3: MUL TEMP[3], IN[2].zzzz, TEMP[2]
>  4: MAX TEMP[3], TEMP[3], IMM[0].wwww
>  5: TEX TEMP[2].x, IN[3], SAMP[2], 2D
>  6: MUL TEMP[3], TEMP[3], TEMP[2].xxxx
>  7: MUL TEMP[4], TEMP[0], IN[0]
>  8: MUL TEMP[5], TEMP[4], CONST[1]
>  9: MUL TEMP[6], TEMP[3], TEMP[5]
> 10: MAD TEMP[6], TEMP[4], CONST[2], TEMP[6]
> 11: MOV OUT[0].w, IMM[0].wwww
> 12: LRP OUT[0].xyz, IN[1].xxxx, CONST[3], TEMP[6]
> 13: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: TEX temp[0], input[5], 2D[0];
>  1: MOV temp[1], temp[2];
>  2: TXP temp[2], input[4], 2D[1];
>  3: MUL temp[3], input[2].zzzz, temp[2];
>  4: MAX temp[3], temp[3], const[4].wwww;
>  5: TEX temp[2].x, input[3], 2D[2];
>  6: MUL temp[3], temp[3], temp[2].xxxx;
>  7: MUL temp[4], temp[0], input[0];
>  8: MUL temp[5], temp[4], const[1];
>  9: MUL temp[6], temp[3], temp[5];
> 10: MAD temp[6], temp[4], const[2], temp[6];
> 11: MOV output[0].w, const[4].wwww;
> 12: LRP output[0].xyz, input[1].xxxx, const[3], temp[6];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: TEX temp[0], input[5], 2D[0];
>  1: MOV temp[1], temp[2];
>  2: TXP temp[2], input[4], 2D[1];
>  3: MUL temp[3], input[2].zzzz, temp[2];
>  4: MAX temp[3], temp[3], const[4].wwww;
>  5: TEX temp[2].x, input[3], 2D[2];
>  6: MUL temp[3], temp[3], temp[2].xxxx;
>  7: MUL temp[4], temp[0], input[0];
>  8: MUL temp[5], temp[4], const[1];
>  9: MUL temp[6], temp[3], temp[5];
> 10: MAD temp[6], temp[4], const[2], temp[6];
> 11: MOV output[0].w, const[4].wwww;
> 12: LRP output[0].xyz, input[1].xxxx, const[3], temp[6];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: TEX temp[0], input[5], 2D[0];
>  1: MOV temp[1], temp[2];
>  2: TXP temp[2], input[4], 2D[1];
>  3: MUL temp[3], input[2].zzzz, temp[2];
>  4: MAX temp[3], temp[3], const[4].wwww;
>  5: TEX temp[2].x, input[3], 2D[2];
>  6: MUL temp[3], temp[3], temp[2].xxxx;
>  7: MUL temp[4], temp[0], input[0];
>  8: MUL temp[5], temp[4], const[1];
>  9: MUL temp[6], temp[3], temp[5];
> 10: MAD temp[6], temp[4], const[2], temp[6];
> 11: MOV output[0].w, const[4].wwww;
> 12: LRP output[0].xyz, input[1].xxxx, const[3], temp[6];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: TEX temp[0], input[5], 2D[0];
>  1: MOV temp[1], temp[2];
>  2: TXP temp[2], input[4], 2D[1];
>  3: MUL temp[3], input[2].zzzz, temp[2];
>  4: MAX temp[3], temp[3], const[4].wwww;
>  5: TEX temp[2].x, input[3], 2D[2];
>  6: MUL temp[3], temp[3], temp[2].xxxx;
>  7: MUL temp[4], temp[0], input[0];
>  8: MUL temp[5], temp[4], const[1];
>  9: MUL temp[6], temp[3], temp[5];
> 10: MAD temp[6], temp[4], const[2], temp[6];
> 11: MOV output[0].w, const[4].wwww;
> 12: ADD temp[7], const[3], -temp[6];
> 13: MAD output[0].xyz, input[1].xxxx, temp[7], temp[6];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: TEX temp[0], input[5], 2D[0];
>  1: MOV temp[1], temp[2];
>  2: TXP temp[2], input[4], 2D[1];
>  3: MUL temp[3], input[2].zzzz, temp[2];
>  4: MAX temp[3], temp[3], const[4].wwww;
>  5: TEX temp[8], input[3], 2D[2];
>  6: MOV temp[2].x, temp[8];
>  7: MUL temp[3], temp[3], temp[2].xxxx;
>  8: MUL temp[4], temp[0], input[0];
>  9: MUL temp[5], temp[4], const[1];
> 10: MUL temp[6], temp[3], temp[5];
> 11: MAD temp[6], temp[4], const[2], temp[6];
> 12: MOV output[0].w, const[4].wwww;
> 13: ADD temp[7], const[3], -temp[6];
> 14: MAD output[0].xyz, input[1].xxxx, temp[7], temp[6];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: TEX temp[0].xyz, input[5].xy__, 2D[0];
>  1: TXP temp[2].xyz, input[4].xy_w, 2D[1];
>  2: MUL temp[3].xyz, input[2].zzz_, temp[2].xyz_;
>  3: MAX temp[3].xyz, temp[3].xyz_, const[4].www_;
>  4: TEX temp[8].x, input[3].xy__, 2D[2];
>  5: MOV temp[2].x, temp[8].x___;
>  6: MUL temp[3].xyz, temp[3].xyz_, temp[2].xxx_;
>  7: MUL temp[4].xyz, temp[0].xyz_, input[0].xyz_;
>  8: MUL temp[5].xyz, temp[4].xyz_, const[1].xyz_;
>  9: MUL temp[6].xyz, temp[3].xyz_, temp[5].xyz_;
> 10: MAD temp[6].xyz, temp[4].xyz_, const[2].xyz_, temp[6].xyz_;
> 11: MOV output[0].w, const[4].___w;
> 12: ADD temp[7].xyz, const[3].xyz_, -temp[6].xyz_;
> 13: MAD output[0].xyz, input[1].xxx_, temp[7].xyz_, temp[6].xyz_;
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: TEX temp[0].xyz, input[5].xy__, 2D[0];
>  1: TXP temp[2].xyz, input[4].xy_w, 2D[1];
>  2: MUL temp[3].xyz, input[2].zzz_, temp[2].xyz_;
>  3: MAX temp[3].xyz, temp[3].xyz_, none.000_;
>  4: TEX temp[8].x, input[3].xy__, 2D[2];
>  5: MUL temp[3].xyz, temp[3].xyz_, temp[8].xxx_;
>  6: MUL temp[4].xyz, temp[0].xyz_, input[0].xyz_;
>  7: MUL temp[5].xyz, temp[4].xyz_, const[1].xyz_;
>  8: MUL temp[6].xyz, temp[3].xyz_, temp[5].xyz_;
>  9: MAD temp[6].xyz, temp[4].xyz_, const[2].xyz_, temp[6].xyz_;
> 10: MOV output[0].w, none.___0;
> 11: ADD temp[7].xyz, const[3].xyz_, -temp[6].xyz_;
> 12: MAD output[0].xyz, input[1].xxx_, temp[7].xyz_, temp[6].xyz_;
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: TEX temp[0].xyz, input[5].xy__, 2D[0];
>  1: TXP temp[2].xyz, input[4].xy_w, 2D[1];
>  2: MUL temp[3].xyz, input[2].zzz_, temp[2].xyz_;
>  3: MAX temp[3].xyz, temp[3].xyz_, none.000_;
>  4: TEX temp[8].x, input[3].xy__, 2D[2];
>  5: MUL temp[3].xyz, temp[3].xyz_, temp[8].xxx_;
>  6: MUL temp[4].xyz, temp[0].xyz_, input[0].xyz_;
>  7: MUL temp[5].xyz, temp[4].xyz_, const[1].xyz_;
>  8: MUL temp[6].xyz, temp[3].xyz_, temp[5].xyz_;
>  9: MAD temp[6].xyz, temp[4].xyz_, const[2].xyz_, temp[6].xyz_;
> 10: MOV output[0].w, none.___0;
> 11: ADD temp[7].xyz, const[3].xyz_, -temp[6].xyz_;
> 12: MAD output[0].xyz, input[1].xxx_, temp[7].xyz_, temp[6].xyz_;
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: TEX temp[0].xyz, input[5].xy__, 2D[0];
>  1: TXP temp[2].xyz, input[4].xy_w, 2D[1];
>  2: src0.xyz = input[2], src1.xyz = temp[2]
>     MAD temp[3].xyz, src0.zzz, src1.xyz, src0.000
>  3: src0.xyz = temp[3]
>     MAX temp[3].xyz, src0.xyz, src0.000
>  4: TEX temp[8].x, input[3].xy__, 2D[2];
>  5: src0.xyz = temp[3], src1.xyz = temp[8]
>     MAD temp[3].xyz, src0.xyz, src1.xxx, src0.000
>  6: src0.xyz = temp[0], src1.xyz = input[0]
>     MAD temp[4].xyz, src0.xyz, src1.xyz, src0.000
>  7: src0.xyz = temp[4], src1.xyz = const[1]
>     MAD temp[5].xyz, src0.xyz, src1.xyz, src0.000
>  8: src0.xyz = temp[3], src1.xyz = temp[5]
>     MAD temp[6].xyz, src0.xyz, src1.xyz, src0.000
>  9: src0.xyz = temp[4], src1.xyz = const[2], src2.xyz = temp[6]
>     MAD temp[6].xyz, src0.xyz, src1.xyz, src2.xyz
> 10: 
>     MAD color[0].w, src0.0, src0.1, src0.0
> 11: src0.xyz = const[3], src1.xyz = temp[6]
>     MAD temp[7].xyz, src0.xyz, src0.111, -src1.xyz
> 12: src0.xyz = input[1], src1.xyz = temp[7], src2.xyz = temp[6]
>     MAD color[0].xyz, src0.xxx, src1.xyz, src2.xyz
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[8].x, input[3].xy__, 2D[2];
>  2: TXP temp[2].xyz, input[4].xy_w, 2D[1];
>  3: TEX temp[0].xyz, input[5].xy__, 2D[0];
>  4: src0.xyz = temp[0], src1.xyz = input[0]
>     MAD temp[4].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.0, src0.1, src0.0
>  5: src0.xyz = temp[4], src1.xyz = const[1]
>     MAD temp[5].xyz, src0.xyz, src1.xyz, src0.000
>  6: src0.xyz = input[2], src1.xyz = temp[2]
>     MAD temp[3].xyz, src0.zzz, src1.xyz, src0.000
>  7: src0.xyz = temp[3]
>     MAX temp[3].xyz, src0.xyz, src0.000
>  8: src0.xyz = temp[3], src1.xyz = temp[8]
>     MAD temp[3].xyz, src0.xyz, src1.xxx, src0.000
>  9: src0.xyz = temp[3], src1.xyz = temp[5]
>     MAD temp[6].xyz, src0.xyz, src1.xyz, src0.000
> 10: src0.xyz = temp[4], src1.xyz = const[2], src2.xyz = temp[6]
>     MAD temp[6].xyz, src0.xyz, src1.xyz, src2.xyz
> 11: src0.xyz = const[3], src1.xyz = temp[6]
>     MAD temp[7].xyz, src0.xyz, src0.111, -src1.xyz
> 12: src0.xyz = input[1], src1.xyz = temp[7], src2.xyz = temp[6]
>     MAD color[0].xyz, src0.xxx, src1.xyz, src2.xyz
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[6].x, temp[3].xy__, 2D[2];
>  2: TXP temp[4].xyz, temp[4].xy_w, 2D[1];
>  3: TEX temp[3].xyz, temp[5].xy__, 2D[0];
>  4: src0.xyz = temp[3], src1.xyz = temp[0]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.0, src0.1, src0.0
>  5: src0.xyz = temp[3], src1.xyz = const[1]
>     MAD temp[5].xyz, src0.xyz, src1.xyz, src0.000
>  6: src0.xyz = temp[2], src1.xyz = temp[4]
>     MAD temp[0].xyz, src0.zzz, src1.xyz, src0.000
>  7: src0.xyz = temp[0]
>     MAX temp[0].xyz, src0.xyz, src0.000
>  8: src0.xyz = temp[0], src1.xyz = temp[6]
>     MAD temp[0].xyz, src0.xyz, src1.xxx, src0.000
>  9: src0.xyz = temp[0], src1.xyz = temp[5]
>     MAD temp[0].xyz, src0.xyz, src1.xyz, src0.000
> 10: src0.xyz = temp[3], src1.xyz = const[2], src2.xyz = temp[0]
>     MAD temp[0].xyz, src0.xyz, src1.xyz, src2.xyz
> 11: src0.xyz = const[3], src1.xyz = temp[0]
>     MAD temp[2].xyz, src0.xyz, src0.111, -src1.xyz
> 12: src0.xyz = temp[1], src1.xyz = temp[2], src2.xyz = temp[0]
>     MAD color[0].xyz, src0.xxx, src1.xyz, src2.xyz
>pc=28*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 8, tex_end: 2  (code_addr: 00440200)
>  TEX:
>    TEX t6, t3, texture[2]   (00009183)
>    TXP t4, t4, texture[1]   (00018904)
>    TEX t3, t5, texture[0]   (000080c5)
>  0: xyz:  t3  t0  t0 -> t3.xyz               (038c0003)
>       w:  t0  t0  t0 -> o0.w                 (01000000)
>     xyz:   t3.xyz   t0.xyz      0.0    op: 00050200
>       w:      0.0      1.0      0.0    op: 00040890
>  1: xyz:  t3  c1  t0 -> t5.xyz               (03940843)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t3.xyz   c1.xyz      0.0    op: 00050200
>       w:     t3.x     t3.x     t3.x    op: 00000000
>  2: xyz:  t2  t4  t0 -> t0.xyz               (03800102)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.zzz   t4.xyz      0.0    op: 00050203
>       w:     t2.x     t2.x     t2.x    op: 00000000
>  3: xyz:  t0  t0  t0 -> t0.xyz               (03800000)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xyz      0.0   t0.xxx    op: 02804a00
>       w:     t0.x     t0.x     t0.x    op: 00000000
>  4: xyz:  t0  t6  t0 -> t0.xyz               (03800180)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xyz   t6.xxx      0.0    op: 00050280
>       w:     t0.x     t0.x     t0.x    op: 00000000
>  5: xyz:  t0  t5  t0 -> t0.xyz               (03800140)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xyz   t5.xyz      0.0    op: 00050200
>       w:     t0.x     t0.x     t0.x    op: 00000000
>  6: xyz:  t3  c2  t0 -> t0.xyz               (03800883)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t3.xyz   c2.xyz   t0.xyz    op: 00020200
>       w:     t3.x     t3.x     t3.x    op: 00000000
>  7: xyz:  c3  t0  t0 -> t2.xyz               (03880023)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   c3.xyz      1.0  -t0.xyz    op: 00090a80
>       w:     c3.x     c3.x     c3.x    op: 00000000
>  8: xyz:  t1  t2  t0 -> o0.xyz               (1c000081)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xxx   t2.xyz   t0.xyz    op: 00020201
>       w:     t1.x     t1.x     t1.x    op: 00000000
>r300: Initial fragment program
>FRAG
>DCL IN[0], COLOR, LINEAR
>DCL IN[1], COLOR[1], LINEAR
>DCL IN[2], GENERIC[0], PERSPECTIVE
>DCL IN[3], GENERIC[2], PERSPECTIVE
>DCL IN[4], GENERIC[3], PERSPECTIVE
>DCL IN[5], GENERIC[4], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL SAMP[1]
>DCL SAMP[2]
>DCL CONST[0]
>DCL CONST[2..4]
>DCL TEMP[0..6]
>IMM FLT32 {    0.3000,     0.5900,     0.1100,     0.0000}
>  0: TEX TEMP[0], IN[5], SAMP[0], 2D
>  1: MOV TEMP[1], TEMP[2]
>  2: SUB TEMP[2].w, TEMP[0].wwww, CONST[0]
>  3: KIL TEMP[2].wwww
>  4: TXP TEMP[2], IN[4], SAMP[1], 2D
>  5: MUL TEMP[3], IN[2].zzzz, TEMP[2]
>  6: MAX TEMP[3], TEMP[3], IMM[0].wwww
>  7: TEX TEMP[2].x, IN[3], SAMP[2], 2D
>  8: MUL TEMP[3], TEMP[3], TEMP[2].xxxx
>  9: MUL TEMP[4], TEMP[0], IN[0]
> 10: MUL TEMP[5], TEMP[4], CONST[2]
> 11: MUL TEMP[6], TEMP[3], TEMP[5]
> 12: MAD TEMP[6], TEMP[4], CONST[3], TEMP[6]
> 13: MOV OUT[0].w, IMM[0].wwww
> 14: LRP OUT[0].xyz, IN[1].xxxx, CONST[4], TEMP[6]
> 15: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: TEX temp[0], input[5], 2D[0];
>  1: MOV temp[1], temp[2];
>  2: SUB temp[2].w, temp[0].wwww, const[0];
>  3: KIL temp[2].wwww;
>  4: TXP temp[2], input[4], 2D[1];
>  5: MUL temp[3], input[2].zzzz, temp[2];
>  6: MAX temp[3], temp[3], const[5].wwww;
>  7: TEX temp[2].x, input[3], 2D[2];
>  8: MUL temp[3], temp[3], temp[2].xxxx;
>  9: MUL temp[4], temp[0], input[0];
> 10: MUL temp[5], temp[4], const[2];
> 11: MUL temp[6], temp[3], temp[5];
> 12: MAD temp[6], temp[4], const[3], temp[6];
> 13: MOV output[0].w, const[5].wwww;
> 14: LRP output[0].xyz, input[1].xxxx, const[4], temp[6];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: TEX temp[0], input[5], 2D[0];
>  1: MOV temp[1], temp[2];
>  2: SUB temp[2].w, temp[0].wwww, const[0];
>  3: KIL temp[2].wwww;
>  4: TXP temp[2], input[4], 2D[1];
>  5: MUL temp[3], input[2].zzzz, temp[2];
>  6: MAX temp[3], temp[3], const[5].wwww;
>  7: TEX temp[2].x, input[3], 2D[2];
>  8: MUL temp[3], temp[3], temp[2].xxxx;
>  9: MUL temp[4], temp[0], input[0];
> 10: MUL temp[5], temp[4], const[2];
> 11: MUL temp[6], temp[3], temp[5];
> 12: MAD temp[6], temp[4], const[3], temp[6];
> 13: MOV output[0].w, const[5].wwww;
> 14: LRP output[0].xyz, input[1].xxxx, const[4], temp[6];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: TEX temp[0], input[5], 2D[0];
>  1: MOV temp[1], temp[2];
>  2: SUB temp[2].w, temp[0].wwww, const[0];
>  3: KIL temp[2].wwww;
>  4: TXP temp[2], input[4], 2D[1];
>  5: MUL temp[3], input[2].zzzz, temp[2];
>  6: MAX temp[3], temp[3], const[5].wwww;
>  7: TEX temp[2].x, input[3], 2D[2];
>  8: MUL temp[3], temp[3], temp[2].xxxx;
>  9: MUL temp[4], temp[0], input[0];
> 10: MUL temp[5], temp[4], const[2];
> 11: MUL temp[6], temp[3], temp[5];
> 12: MAD temp[6], temp[4], const[3], temp[6];
> 13: MOV output[0].w, const[5].wwww;
> 14: LRP output[0].xyz, input[1].xxxx, const[4], temp[6];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: TEX temp[0], input[5], 2D[0];
>  1: MOV temp[1], temp[2];
>  2: ADD temp[2].w, temp[0].wwww, -const[0];
>  3: KIL temp[2].wwww;
>  4: TXP temp[2], input[4], 2D[1];
>  5: MUL temp[3], input[2].zzzz, temp[2];
>  6: MAX temp[3], temp[3], const[5].wwww;
>  7: TEX temp[2].x, input[3], 2D[2];
>  8: MUL temp[3], temp[3], temp[2].xxxx;
>  9: MUL temp[4], temp[0], input[0];
> 10: MUL temp[5], temp[4], const[2];
> 11: MUL temp[6], temp[3], temp[5];
> 12: MAD temp[6], temp[4], const[3], temp[6];
> 13: MOV output[0].w, const[5].wwww;
> 14: ADD temp[7], const[4], -temp[6];
> 15: MAD output[0].xyz, input[1].xxxx, temp[7], temp[6];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: TEX temp[0], input[5], 2D[0];
>  1: MOV temp[1], temp[2];
>  2: ADD temp[2].w, temp[0].wwww, -const[0];
>  3: KIL temp[2].wwww;
>  4: TXP temp[2], input[4], 2D[1];
>  5: MUL temp[3], input[2].zzzz, temp[2];
>  6: MAX temp[3], temp[3], const[5].wwww;
>  7: TEX temp[8], input[3], 2D[2];
>  8: MOV temp[2].x, temp[8];
>  9: MUL temp[3], temp[3], temp[2].xxxx;
> 10: MUL temp[4], temp[0], input[0];
> 11: MUL temp[5], temp[4], const[2];
> 12: MUL temp[6], temp[3], temp[5];
> 13: MAD temp[6], temp[4], const[3], temp[6];
> 14: MOV output[0].w, const[5].wwww;
> 15: ADD temp[7], const[4], -temp[6];
> 16: MAD output[0].xyz, input[1].xxxx, temp[7], temp[6];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: TEX temp[0], input[5].xy__, 2D[0];
>  1: ADD temp[2].w, temp[0].___w, -const[0].___w;
>  2: KIL temp[2].wwww;
>  3: TXP temp[2].xyz, input[4].xy_w, 2D[1];
>  4: MUL temp[3].xyz, input[2].zzz_, temp[2].xyz_;
>  5: MAX temp[3].xyz, temp[3].xyz_, const[5].www_;
>  6: TEX temp[8].x, input[3].xy__, 2D[2];
>  7: MOV temp[2].x, temp[8].x___;
>  8: MUL temp[3].xyz, temp[3].xyz_, temp[2].xxx_;
>  9: MUL temp[4].xyz, temp[0].xyz_, input[0].xyz_;
> 10: MUL temp[5].xyz, temp[4].xyz_, const[2].xyz_;
> 11: MUL temp[6].xyz, temp[3].xyz_, temp[5].xyz_;
> 12: MAD temp[6].xyz, temp[4].xyz_, const[3].xyz_, temp[6].xyz_;
> 13: MOV output[0].w, const[5].___w;
> 14: ADD temp[7].xyz, const[4].xyz_, -temp[6].xyz_;
> 15: MAD output[0].xyz, input[1].xxx_, temp[7].xyz_, temp[6].xyz_;
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: TEX temp[0], input[5].xy__, 2D[0];
>  1: ADD temp[2].w, temp[0].___w, -const[0].___w;
>  2: KIL temp[2].wwww;
>  3: TXP temp[2].xyz, input[4].xy_w, 2D[1];
>  4: MUL temp[3].xyz, input[2].zzz_, temp[2].xyz_;
>  5: MAX temp[3].xyz, temp[3].xyz_, none.000_;
>  6: TEX temp[8].x, input[3].xy__, 2D[2];
>  7: MUL temp[3].xyz, temp[3].xyz_, temp[8].xxx_;
>  8: MUL temp[4].xyz, temp[0].xyz_, input[0].xyz_;
>  9: MUL temp[5].xyz, temp[4].xyz_, const[2].xyz_;
> 10: MUL temp[6].xyz, temp[3].xyz_, temp[5].xyz_;
> 11: MAD temp[6].xyz, temp[4].xyz_, const[3].xyz_, temp[6].xyz_;
> 12: MOV output[0].w, none.___0;
> 13: ADD temp[7].xyz, const[4].xyz_, -temp[6].xyz_;
> 14: MAD output[0].xyz, input[1].xxx_, temp[7].xyz_, temp[6].xyz_;
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: TEX temp[0], input[5].xy__, 2D[0];
>  1: ADD temp[2].w, temp[0].___w, -const[0].___w;
>  2: MOV temp[1], temp[2].wwww;
>  3: KIL temp[1];
>  4: TXP temp[2].xyz, input[4].xy_w, 2D[1];
>  5: MUL temp[3].xyz, input[2].zzz_, temp[2].xyz_;
>  6: MAX temp[3].xyz, temp[3].xyz_, none.000_;
>  7: TEX temp[8].x, input[3].xy__, 2D[2];
>  8: MUL temp[3].xyz, temp[3].xyz_, temp[8].xxx_;
>  9: MUL temp[4].xyz, temp[0].xyz_, input[0].xyz_;
> 10: MUL temp[5].xyz, temp[4].xyz_, const[2].xyz_;
> 11: MUL temp[6].xyz, temp[3].xyz_, temp[5].xyz_;
> 12: MAD temp[6].xyz, temp[4].xyz_, const[3].xyz_, temp[6].xyz_;
> 13: MOV output[0].w, none.___0;
> 14: ADD temp[7].xyz, const[4].xyz_, -temp[6].xyz_;
> 15: MAD output[0].xyz, input[1].xxx_, temp[7].xyz_, temp[6].xyz_;
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: TEX temp[0], input[5].xy__, 2D[0];
>  1: src0.w = temp[0], src1.w = const[0]
>     MAD temp[2].w, src0.w, src0.1, -src1.w
>  2: src0.w = temp[2]
>     MAD temp[1].xyz, src0.www, src0.111, src0.000
>     MAD temp[1].w, src0.w, src0.1, src0.0
>  3: KIL temp[1];
>  4: TXP temp[2].xyz, input[4].xy_w, 2D[1];
>  5: src0.xyz = input[2], src1.xyz = temp[2]
>     MAD temp[3].xyz, src0.zzz, src1.xyz, src0.000
>  6: src0.xyz = temp[3]
>     MAX temp[3].xyz, src0.xyz, src0.000
>  7: TEX temp[8].x, input[3].xy__, 2D[2];
>  8: src0.xyz = temp[3], src1.xyz = temp[8]
>     MAD temp[3].xyz, src0.xyz, src1.xxx, src0.000
>  9: src0.xyz = temp[0], src1.xyz = input[0]
>     MAD temp[4].xyz, src0.xyz, src1.xyz, src0.000
> 10: src0.xyz = temp[4], src1.xyz = const[2]
>     MAD temp[5].xyz, src0.xyz, src1.xyz, src0.000
> 11: src0.xyz = temp[3], src1.xyz = temp[5]
>     MAD temp[6].xyz, src0.xyz, src1.xyz, src0.000
> 12: src0.xyz = temp[4], src1.xyz = const[3], src2.xyz = temp[6]
>     MAD temp[6].xyz, src0.xyz, src1.xyz, src2.xyz
> 13: 
>     MAD color[0].w, src0.0, src0.1, src0.0
> 14: src0.xyz = const[4], src1.xyz = temp[6]
>     MAD temp[7].xyz, src0.xyz, src0.111, -src1.xyz
> 15: src0.xyz = input[1], src1.xyz = temp[7], src2.xyz = temp[6]
>     MAD color[0].xyz, src0.xxx, src1.xyz, src2.xyz
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[8].x, input[3].xy__, 2D[2];
>  2: TXP temp[2].xyz, input[4].xy_w, 2D[1];
>  3: TEX temp[0], input[5].xy__, 2D[0];
>  4: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = input[0], src1.w = const[0]
>     MAD temp[4].xyz, src0.xyz, src1.xyz, src0.000
>     MAD temp[2].w, src0.w, src0.1, -src1.w
>  5: src0.w = temp[2]
>     MAD temp[1].xyz, src0.www, src0.111, src0.000
>     MAD temp[1].w, src0.w, src0.1, src0.0
>  6: src0.xyz = temp[4], src1.xyz = const[2]
>     MAD temp[5].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.0, src0.1, src0.0
>  7: src0.xyz = input[2], src1.xyz = temp[2]
>     MAD temp[3].xyz, src0.zzz, src1.xyz, src0.000
>  8: src0.xyz = temp[3]
>     MAX temp[3].xyz, src0.xyz, src0.000
>  9: src0.xyz = temp[3], src1.xyz = temp[8]
>     MAD temp[3].xyz, src0.xyz, src1.xxx, src0.000
> 10: src0.xyz = temp[3], src1.xyz = temp[5]
>     MAD temp[6].xyz, src0.xyz, src1.xyz, src0.000
> 11: src0.xyz = temp[4], src1.xyz = const[3], src2.xyz = temp[6]
>     MAD temp[6].xyz, src0.xyz, src1.xyz, src2.xyz
> 12: src0.xyz = const[4], src1.xyz = temp[6]
>     MAD temp[7].xyz, src0.xyz, src0.111, -src1.xyz
> 13: src0.xyz = input[1], src1.xyz = temp[7], src2.xyz = temp[6]
>     MAD color[0].xyz, src0.xxx, src1.xyz, src2.xyz
> 14: BEGIN_TEX;
> 15: KIL temp[1];
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[6].x, temp[3].xy__, 2D[2];
>  2: TXP temp[4].xyz, temp[4].xy_w, 2D[1];
>  3: TEX temp[3], temp[5].xy__, 2D[0];
>  4: src0.xyz = temp[3], src0.w = temp[3], src1.xyz = temp[0], src1.w = const[0]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src0.000
>     MAD temp[4].w, src0.w, src0.1, -src1.w
>  5: src0.w = temp[4]
>     MAD temp[0].xyz, src0.www, src0.111, src0.000
>     MAD temp[0].w, src0.w, src0.1, src0.0
>  6: src0.xyz = temp[3], src1.xyz = const[2]
>     MAD temp[5].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.0, src0.1, src0.0
>  7: src0.xyz = temp[2], src1.xyz = temp[4]
>     MAD temp[2].xyz, src0.zzz, src1.xyz, src0.000
>  8: src0.xyz = temp[2]
>     MAX temp[2].xyz, src0.xyz, src0.000
>  9: src0.xyz = temp[2], src1.xyz = temp[6]
>     MAD temp[2].xyz, src0.xyz, src1.xxx, src0.000
> 10: src0.xyz = temp[2], src1.xyz = temp[5]
>     MAD temp[2].xyz, src0.xyz, src1.xyz, src0.000
> 11: src0.xyz = temp[3], src1.xyz = const[3], src2.xyz = temp[2]
>     MAD temp[2].xyz, src0.xyz, src1.xyz, src2.xyz
> 12: src0.xyz = const[4], src1.xyz = temp[2]
>     MAD temp[3].xyz, src0.xyz, src0.111, -src1.xyz
> 13: src0.xyz = temp[1], src1.xyz = temp[3], src2.xyz = temp[2]
>     MAD color[0].xyz, src0.xxx, src1.xyz, src2.xyz
> 14: BEGIN_TEX;
> 15: KIL temp[0];
>pc=29*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 9, tex_end: 2  (code_addr: 00440240)
>  TEX:
>    TEX t6, t3, texture[2]   (00009183)
>    TXP t4, t4, texture[1]   (00018904)
>    TEX t3, t5, texture[0]   (000080c5)
>  0: xyz:  t3  t0  t0 -> t3.xyz               (038c0003)
>       w:  t3  c0  t0 -> t4.w                 (00900803)
>     xyz:   t3.xyz   t0.xyz      0.0    op: 00050200
>       w:     t3.w      1.0    -c0.w    op: 000a8889
>  1: xyz:  t0  t0  t0 -> t0.xyz               (03800000)
>       w:  t4  t0  t0 -> t0.w                 (00800004)
>     xyz:   t4.www      1.0      0.0    op: 00050a8c
>       w:     t4.w      1.0      0.0    op: 00040889
>  2: xyz:  t3  c2  t0 -> t5.xyz               (03940883)
>       w:  t0  t0  t0 -> o0.w                 (01000000)
>     xyz:   t3.xyz   c2.xyz      0.0    op: 00050200
>       w:      0.0      1.0      0.0    op: 00040890
>  3: xyz:  t2  t4  t0 -> t2.xyz               (03880102)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.zzz   t4.xyz      0.0    op: 00050203
>       w:     t2.x     t2.x     t2.x    op: 00000000
>  4: xyz:  t2  t0  t0 -> t2.xyz               (03880002)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz      0.0   t2.xxx    op: 02804a00
>       w:     t2.x     t2.x     t2.x    op: 00000000
>  5: xyz:  t2  t6  t0 -> t2.xyz               (03880182)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz   t6.xxx      0.0    op: 00050280
>       w:     t2.x     t2.x     t2.x    op: 00000000
>  6: xyz:  t2  t5  t0 -> t2.xyz               (03880142)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz   t5.xyz      0.0    op: 00050200
>       w:     t2.x     t2.x     t2.x    op: 00000000
>  7: xyz:  t3  c3  t2 -> t2.xyz               (038828c3)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t3.xyz   c3.xyz   t2.xyz    op: 00020200
>       w:     t3.x     t3.x     t3.x    op: 00000000
>  8: xyz:  c4  t2  t0 -> t3.xyz               (038c00a4)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   c4.xyz      1.0  -t2.xyz    op: 00090a80
>       w:     c4.x     c4.x     c4.x    op: 00000000
>  9: xyz:  t1  t3  t2 -> o0.xyz               (1c0020c1)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xxx   t3.xyz   t2.xyz    op: 00020201
>       w:     t1.x     t1.x     t1.x    op: 00000000
>NODE 1: alu_offset: 10, tex_offset: 3, alu_end: 0, tex_end: 0  (code_addr: 0000300a)
>  TEX:
>    KIL t0, t0, texture[0]   (00010000)
> 10: xyz:  t0  t0  t0 ->                      (00000000)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xxx   t0.xxx   t0.xxx    op: 00004081
>       w:     t0.x     t0.x     t0.x    op: 00000000
>r300: Initial fragment program
>FRAG
>DCL IN[0], COLOR, LINEAR
>DCL IN[1], COLOR[1], LINEAR
>DCL IN[2], GENERIC[0], PERSPECTIVE
>DCL IN[3], GENERIC[2], PERSPECTIVE
>DCL IN[4], GENERIC[3], PERSPECTIVE
>DCL IN[5], GENERIC[4], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL SAMP[1]
>DCL SAMP[2]
>DCL SAMP[3]
>DCL CONST[1..3]
>DCL TEMP[0..6]
>IMM FLT32 {    0.3000,     0.5900,     0.1100,     0.0000}
>  0: TEX TEMP[0], IN[5], SAMP[0], 2D
>  1: MOV TEMP[1], TEMP[2]
>  2: TXP TEMP[2], IN[4], SAMP[1], 2D
>  3: MUL TEMP[3], IN[2].zzzz, TEMP[2]
>  4: MAX TEMP[3], TEMP[3], IMM[0].wwww
>  5: TEX TEMP[2].x, IN[3], SAMP[2], 2D
>  6: MUL TEMP[3], TEMP[3], TEMP[2].xxxx
>  7: MUL TEMP[4], TEMP[0], IN[0]
>  8: MUL TEMP[5], TEMP[4], CONST[1]
>  9: MUL TEMP[6], TEMP[3], TEMP[5]
> 10: MAD TEMP[6], TEMP[4], CONST[2], TEMP[6]
> 11: TEX TEMP[2], IN[5], SAMP[3], 2D
> 12: ADD TEMP[6], TEMP[2], TEMP[6]
> 13: MOV OUT[0].w, IMM[0].wwww
> 14: LRP OUT[0].xyz, IN[1].xxxx, CONST[3], TEMP[6]
> 15: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: TEX temp[0], input[5], 2D[0];
>  1: MOV temp[1], temp[2];
>  2: TXP temp[2], input[4], 2D[1];
>  3: MUL temp[3], input[2].zzzz, temp[2];
>  4: MAX temp[3], temp[3], const[4].wwww;
>  5: TEX temp[2].x, input[3], 2D[2];
>  6: MUL temp[3], temp[3], temp[2].xxxx;
>  7: MUL temp[4], temp[0], input[0];
>  8: MUL temp[5], temp[4], const[1];
>  9: MUL temp[6], temp[3], temp[5];
> 10: MAD temp[6], temp[4], const[2], temp[6];
> 11: TEX temp[2], input[5], 2D[3];
> 12: ADD temp[6], temp[2], temp[6];
> 13: MOV output[0].w, const[4].wwww;
> 14: LRP output[0].xyz, input[1].xxxx, const[3], temp[6];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: TEX temp[0], input[5], 2D[0];
>  1: MOV temp[1], temp[2];
>  2: TXP temp[2], input[4], 2D[1];
>  3: MUL temp[3], input[2].zzzz, temp[2];
>  4: MAX temp[3], temp[3], const[4].wwww;
>  5: TEX temp[2].x, input[3], 2D[2];
>  6: MUL temp[3], temp[3], temp[2].xxxx;
>  7: MUL temp[4], temp[0], input[0];
>  8: MUL temp[5], temp[4], const[1];
>  9: MUL temp[6], temp[3], temp[5];
> 10: MAD temp[6], temp[4], const[2], temp[6];
> 11: TEX temp[2], input[5], 2D[3];
> 12: ADD temp[6], temp[2], temp[6];
> 13: MOV output[0].w, const[4].wwww;
> 14: LRP output[0].xyz, input[1].xxxx, const[3], temp[6];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: TEX temp[0], input[5], 2D[0];
>  1: MOV temp[1], temp[2];
>  2: TXP temp[2], input[4], 2D[1];
>  3: MUL temp[3], input[2].zzzz, temp[2];
>  4: MAX temp[3], temp[3], const[4].wwww;
>  5: TEX temp[2].x, input[3], 2D[2];
>  6: MUL temp[3], temp[3], temp[2].xxxx;
>  7: MUL temp[4], temp[0], input[0];
>  8: MUL temp[5], temp[4], const[1];
>  9: MUL temp[6], temp[3], temp[5];
> 10: MAD temp[6], temp[4], const[2], temp[6];
> 11: TEX temp[2], input[5], 2D[3];
> 12: ADD temp[6], temp[2], temp[6];
> 13: MOV output[0].w, const[4].wwww;
> 14: LRP output[0].xyz, input[1].xxxx, const[3], temp[6];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: TEX temp[0], input[5], 2D[0];
>  1: MOV temp[1], temp[2];
>  2: TXP temp[2], input[4], 2D[1];
>  3: MUL temp[3], input[2].zzzz, temp[2];
>  4: MAX temp[3], temp[3], const[4].wwww;
>  5: TEX temp[2].x, input[3], 2D[2];
>  6: MUL temp[3], temp[3], temp[2].xxxx;
>  7: MUL temp[4], temp[0], input[0];
>  8: MUL temp[5], temp[4], const[1];
>  9: MUL temp[6], temp[3], temp[5];
> 10: MAD temp[6], temp[4], const[2], temp[6];
> 11: TEX temp[2], input[5], 2D[3];
> 12: ADD temp[6], temp[2], temp[6];
> 13: MOV output[0].w, const[4].wwww;
> 14: ADD temp[7], const[3], -temp[6];
> 15: MAD output[0].xyz, input[1].xxxx, temp[7], temp[6];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: TEX temp[0], input[5], 2D[0];
>  1: MOV temp[1], temp[2];
>  2: TXP temp[2], input[4], 2D[1];
>  3: MUL temp[3], input[2].zzzz, temp[2];
>  4: MAX temp[3], temp[3], const[4].wwww;
>  5: TEX temp[8], input[3], 2D[2];
>  6: MOV temp[2].x, temp[8];
>  7: MUL temp[3], temp[3], temp[2].xxxx;
>  8: MUL temp[4], temp[0], input[0];
>  9: MUL temp[5], temp[4], const[1];
> 10: MUL temp[6], temp[3], temp[5];
> 11: MAD temp[6], temp[4], const[2], temp[6];
> 12: TEX temp[2], input[5], 2D[3];
> 13: ADD temp[6], temp[2], temp[6];
> 14: MOV output[0].w, const[4].wwww;
> 15: ADD temp[7], const[3], -temp[6];
> 16: MAD output[0].xyz, input[1].xxxx, temp[7], temp[6];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: TEX temp[0].xyz, input[5].xy__, 2D[0];
>  1: TXP temp[2].xyz, input[4].xy_w, 2D[1];
>  2: MUL temp[3].xyz, input[2].zzz_, temp[2].xyz_;
>  3: MAX temp[3].xyz, temp[3].xyz_, const[4].www_;
>  4: TEX temp[8].x, input[3].xy__, 2D[2];
>  5: MOV temp[2].x, temp[8].x___;
>  6: MUL temp[3].xyz, temp[3].xyz_, temp[2].xxx_;
>  7: MUL temp[4].xyz, temp[0].xyz_, input[0].xyz_;
>  8: MUL temp[5].xyz, temp[4].xyz_, const[1].xyz_;
>  9: MUL temp[6].xyz, temp[3].xyz_, temp[5].xyz_;
> 10: MAD temp[6].xyz, temp[4].xyz_, const[2].xyz_, temp[6].xyz_;
> 11: TEX temp[2].xyz, input[5].xy__, 2D[3];
> 12: ADD temp[6].xyz, temp[2].xyz_, temp[6].xyz_;
> 13: MOV output[0].w, const[4].___w;
> 14: ADD temp[7].xyz, const[3].xyz_, -temp[6].xyz_;
> 15: MAD output[0].xyz, input[1].xxx_, temp[7].xyz_, temp[6].xyz_;
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: TEX temp[0].xyz, input[5].xy__, 2D[0];
>  1: TXP temp[2].xyz, input[4].xy_w, 2D[1];
>  2: MUL temp[3].xyz, input[2].zzz_, temp[2].xyz_;
>  3: MAX temp[3].xyz, temp[3].xyz_, none.000_;
>  4: TEX temp[8].x, input[3].xy__, 2D[2];
>  5: MUL temp[3].xyz, temp[3].xyz_, temp[8].xxx_;
>  6: MUL temp[4].xyz, temp[0].xyz_, input[0].xyz_;
>  7: MUL temp[5].xyz, temp[4].xyz_, const[1].xyz_;
>  8: MUL temp[6].xyz, temp[3].xyz_, temp[5].xyz_;
>  9: MAD temp[6].xyz, temp[4].xyz_, const[2].xyz_, temp[6].xyz_;
> 10: TEX temp[2].xyz, input[5].xy__, 2D[3];
> 11: ADD temp[6].xyz, temp[2].xyz_, temp[6].xyz_;
> 12: MOV output[0].w, none.___0;
> 13: ADD temp[7].xyz, const[3].xyz_, -temp[6].xyz_;
> 14: MAD output[0].xyz, input[1].xxx_, temp[7].xyz_, temp[6].xyz_;
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: TEX temp[0].xyz, input[5].xy__, 2D[0];
>  1: TXP temp[2].xyz, input[4].xy_w, 2D[1];
>  2: MUL temp[3].xyz, input[2].zzz_, temp[2].xyz_;
>  3: MAX temp[3].xyz, temp[3].xyz_, none.000_;
>  4: TEX temp[8].x, input[3].xy__, 2D[2];
>  5: MUL temp[3].xyz, temp[3].xyz_, temp[8].xxx_;
>  6: MUL temp[4].xyz, temp[0].xyz_, input[0].xyz_;
>  7: MUL temp[5].xyz, temp[4].xyz_, const[1].xyz_;
>  8: MUL temp[6].xyz, temp[3].xyz_, temp[5].xyz_;
>  9: MAD temp[6].xyz, temp[4].xyz_, const[2].xyz_, temp[6].xyz_;
> 10: TEX temp[2].xyz, input[5].xy__, 2D[3];
> 11: ADD temp[6].xyz, temp[2].xyz_, temp[6].xyz_;
> 12: MOV output[0].w, none.___0;
> 13: ADD temp[7].xyz, const[3].xyz_, -temp[6].xyz_;
> 14: MAD output[0].xyz, input[1].xxx_, temp[7].xyz_, temp[6].xyz_;
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: TEX temp[0].xyz, input[5].xy__, 2D[0];
>  1: TXP temp[2].xyz, input[4].xy_w, 2D[1];
>  2: src0.xyz = input[2], src1.xyz = temp[2]
>     MAD temp[3].xyz, src0.zzz, src1.xyz, src0.000
>  3: src0.xyz = temp[3]
>     MAX temp[3].xyz, src0.xyz, src0.000
>  4: TEX temp[8].x, input[3].xy__, 2D[2];
>  5: src0.xyz = temp[3], src1.xyz = temp[8]
>     MAD temp[3].xyz, src0.xyz, src1.xxx, src0.000
>  6: src0.xyz = temp[0], src1.xyz = input[0]
>     MAD temp[4].xyz, src0.xyz, src1.xyz, src0.000
>  7: src0.xyz = temp[4], src1.xyz = const[1]
>     MAD temp[5].xyz, src0.xyz, src1.xyz, src0.000
>  8: src0.xyz = temp[3], src1.xyz = temp[5]
>     MAD temp[6].xyz, src0.xyz, src1.xyz, src0.000
>  9: src0.xyz = temp[4], src1.xyz = const[2], src2.xyz = temp[6]
>     MAD temp[6].xyz, src0.xyz, src1.xyz, src2.xyz
> 10: TEX temp[2].xyz, input[5].xy__, 2D[3];
> 11: src0.xyz = temp[2], src1.xyz = temp[6]
>     MAD temp[6].xyz, src0.xyz, src0.111, src1.xyz
> 12: 
>     MAD color[0].w, src0.0, src0.1, src0.0
> 13: src0.xyz = const[3], src1.xyz = temp[6]
>     MAD temp[7].xyz, src0.xyz, src0.111, -src1.xyz
> 14: src0.xyz = input[1], src1.xyz = temp[7], src2.xyz = temp[6]
>     MAD color[0].xyz, src0.xxx, src1.xyz, src2.xyz
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[8].x, input[3].xy__, 2D[2];
>  2: TXP temp[2].xyz, input[4].xy_w, 2D[1];
>  3: TEX temp[0].xyz, input[5].xy__, 2D[0];
>  4: src0.xyz = temp[0], src1.xyz = input[0]
>     MAD temp[4].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.0, src0.1, src0.0
>  5: src0.xyz = temp[4], src1.xyz = const[1]
>     MAD temp[5].xyz, src0.xyz, src1.xyz, src0.000
>  6: src0.xyz = input[2], src1.xyz = temp[2]
>     MAD temp[3].xyz, src0.zzz, src1.xyz, src0.000
>  7: src0.xyz = temp[3]
>     MAX temp[3].xyz, src0.xyz, src0.000
>  8: src0.xyz = temp[3], src1.xyz = temp[8]
>     MAD temp[3].xyz, src0.xyz, src1.xxx, src0.000
>  9: src0.xyz = temp[3], src1.xyz = temp[5]
>     MAD temp[6].xyz, src0.xyz, src1.xyz, src0.000
> 10: src0.xyz = temp[4], src1.xyz = const[2], src2.xyz = temp[6]
>     MAD temp[6].xyz, src0.xyz, src1.xyz, src2.xyz
> 11: BEGIN_TEX;
> 12: TEX temp[2].xyz, input[5].xy__, 2D[3];
> 13: src0.xyz = temp[2], src1.xyz = temp[6]
>     MAD temp[6].xyz, src0.xyz, src0.111, src1.xyz
> 14: src0.xyz = const[3], src1.xyz = temp[6]
>     MAD temp[7].xyz, src0.xyz, src0.111, -src1.xyz
> 15: src0.xyz = input[1], src1.xyz = temp[7], src2.xyz = temp[6]
>     MAD color[0].xyz, src0.xxx, src1.xyz, src2.xyz
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[7].x, temp[3].xy__, 2D[2];
>  2: TXP temp[4].xyz, temp[4].xy_w, 2D[1];
>  3: TEX temp[3].xyz, temp[5].xy__, 2D[0];
>  4: src0.xyz = temp[3], src1.xyz = temp[0]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.0, src0.1, src0.0
>  5: src0.xyz = temp[3], src1.xyz = const[1]
>     MAD temp[6].xyz, src0.xyz, src1.xyz, src0.000
>  6: src0.xyz = temp[2], src1.xyz = temp[4]
>     MAD temp[0].xyz, src0.zzz, src1.xyz, src0.000
>  7: src0.xyz = temp[0]
>     MAX temp[0].xyz, src0.xyz, src0.000
>  8: src0.xyz = temp[0], src1.xyz = temp[7]
>     MAD temp[0].xyz, src0.xyz, src1.xxx, src0.000
>  9: src0.xyz = temp[0], src1.xyz = temp[6]
>     MAD temp[0].xyz, src0.xyz, src1.xyz, src0.000
> 10: src0.xyz = temp[3], src1.xyz = const[2], src2.xyz = temp[0]
>     MAD temp[0].xyz, src0.xyz, src1.xyz, src2.xyz
> 11: BEGIN_TEX;
> 12: TEX temp[4].xyz, temp[5].xy__, 2D[3];
> 13: src0.xyz = temp[4], src1.xyz = temp[0]
>     MAD temp[0].xyz, src0.xyz, src0.111, src1.xyz
> 14: src0.xyz = const[3], src1.xyz = temp[0]
>     MAD temp[2].xyz, src0.xyz, src0.111, -src1.xyz
> 15: src0.xyz = temp[1], src1.xyz = temp[2], src2.xyz = temp[0]
>     MAD color[0].xyz, src0.xxx, src1.xyz, src2.xyz
>pc=30*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 6, tex_end: 2  (code_addr: 00440180)
>  TEX:
>    TEX t7, t3, texture[2]   (000091c3)
>    TXP t4, t4, texture[1]   (00018904)
>    TEX t3, t5, texture[0]   (000080c5)
>  0: xyz:  t3  t0  t0 -> t3.xyz               (038c0003)
>       w:  t0  t0  t0 -> o0.w                 (01000000)
>     xyz:   t3.xyz   t0.xyz      0.0    op: 00050200
>       w:      0.0      1.0      0.0    op: 00040890
>  1: xyz:  t3  c1  t0 -> t6.xyz               (03980843)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t3.xyz   c1.xyz      0.0    op: 00050200
>       w:     t3.x     t3.x     t3.x    op: 00000000
>  2: xyz:  t2  t4  t0 -> t0.xyz               (03800102)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.zzz   t4.xyz      0.0    op: 00050203
>       w:     t2.x     t2.x     t2.x    op: 00000000
>  3: xyz:  t0  t0  t0 -> t0.xyz               (03800000)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xyz      0.0   t0.xxx    op: 02804a00
>       w:     t0.x     t0.x     t0.x    op: 00000000
>  4: xyz:  t0  t7  t0 -> t0.xyz               (038001c0)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xyz   t7.xxx      0.0    op: 00050280
>       w:     t0.x     t0.x     t0.x    op: 00000000
>  5: xyz:  t0  t6  t0 -> t0.xyz               (03800180)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xyz   t6.xyz      0.0    op: 00050200
>       w:     t0.x     t0.x     t0.x    op: 00000000
>  6: xyz:  t3  c2  t0 -> t0.xyz               (03800883)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t3.xyz   c2.xyz   t0.xyz    op: 00020200
>       w:     t3.x     t3.x     t3.x    op: 00000000
>NODE 1: alu_offset: 7, tex_offset: 3, alu_end: 2, tex_end: 0  (code_addr: 00403087)
>  TEX:
>    TEX t4, t5, texture[3]   (00009905)
>  7: xyz:  t4  t0  t0 -> t0.xyz               (03800004)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t4.xyz      1.0   t0.xyz    op: 00010a80
>       w:     t4.x     t4.x     t4.x    op: 00000000
>  8: xyz:  c3  t0  t0 -> t2.xyz               (03880023)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   c3.xyz      1.0  -t0.xyz    op: 00090a80
>       w:     c3.x     c3.x     c3.x    op: 00000000
>  9: xyz:  t1  t2  t0 -> o0.xyz               (1c000081)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xxx   t2.xyz   t0.xyz    op: 00020201
>       w:     t1.x     t1.x     t1.x    op: 00000000
>r300: Initial fragment program
>FRAG
>DCL IN[0], COLOR, LINEAR
>DCL OUT[0], COLOR
>  0: MOV OUT[0], IN[0]
>  1: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: src0.xyz = input[0], src0.w = input[0]
>     MAD color[0].xyz, src0.xyz, src0.111, src0.000
>     MAD color[0].w, src0.w, src0.1, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: src0.xyz = input[0], src0.w = input[0]
>     MAD color[0].xyz, src0.xyz, src0.111, src0.000
>     MAD color[0].w, src0.w, src0.1, src0.0
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: src0.xyz = temp[0], src0.w = temp[0]
>     MAD color[0].xyz, src0.xyz, src0.111, src0.000
>     MAD color[0].w, src0.w, src0.1, src0.0
>pc=31*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 0, tex_end: 0  (code_addr: 00400000)
>  0: xyz:  t0  t0  t0 -> o0.xyz               (1c000000)
>       w:  t0  t0  t0 -> o0.w                 (01000000)
>     xyz:   t0.xyz      1.0      0.0    op: 00050a80
>       w:     t0.w      1.0      0.0    op: 00040889
>r300: Initial fragment program
>FRAG
>DCL IN[0], COLOR, LINEAR
>DCL IN[1], COLOR[1], LINEAR
>DCL IN[2], GENERIC[4], PERSPECTIVE
>DCL IN[3], GENERIC[5], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL SAMP[1]
>DCL CONST[0..3]
>DCL TEMP[0..3]
>  0: TEX TEMP[0], IN[2], SAMP[0], 2D
>  1: MOV TEMP[1], IN[3]
>  2: TEX TEMP[2], TEMP[1], SAMP[1], CUBE
>  3: MUL TEMP[2], TEMP[2], CONST[0]
>  4: MUL TEMP[3], TEMP[0], CONST[1]
>  5: MUL TEMP[3], TEMP[2], TEMP[3]
>  6: MUL TEMP[3], CONST[2].xxxx, TEMP[3]
>  7: LRP TEMP[3].xyz, IN[1].xxxx, CONST[3], TEMP[3]
>  8: MUL OUT[0], TEMP[3], IN[0]
>  9: MOV OUT[0].w, CONST[2].yyyy
> 10: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: TEX temp[0], input[2], 2D[0];
>  1: MOV temp[1], input[3];
>  2: TEX temp[2], temp[1], CUBE[1];
>  3: MUL temp[2], temp[2], const[0];
>  4: MUL temp[3], temp[0], const[1];
>  5: MUL temp[3], temp[2], temp[3];
>  6: MUL temp[3], const[2].xxxx, temp[3];
>  7: LRP temp[3].xyz, input[1].xxxx, const[3], temp[3];
>  8: MUL output[0], temp[3], input[0];
>  9: MOV output[0].w, const[2].yyyy;
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: TEX temp[0], input[2], 2D[0];
>  1: MOV temp[1], input[3];
>  2: TEX temp[2], temp[1], CUBE[1];
>  3: MUL temp[2], temp[2], const[0];
>  4: MUL temp[3], temp[0], const[1];
>  5: MUL temp[3], temp[2], temp[3];
>  6: MUL temp[3], const[2].xxxx, temp[3];
>  7: LRP temp[3].xyz, input[1].xxxx, const[3], temp[3];
>  8: MUL output[0], temp[3], input[0];
>  9: MOV output[0].w, const[2].yyyy;
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: TEX temp[0], input[2], 2D[0];
>  1: MOV temp[1], input[3];
>  2: TEX temp[2], temp[1], CUBE[1];
>  3: MUL temp[2], temp[2], const[0];
>  4: MUL temp[3], temp[0], const[1];
>  5: MUL temp[3], temp[2], temp[3];
>  6: MUL temp[3], const[2].xxxx, temp[3];
>  7: LRP temp[3].xyz, input[1].xxxx, const[3], temp[3];
>  8: MUL output[0], temp[3], input[0];
>  9: MOV output[0].w, const[2].yyyy;
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: TEX temp[0], input[2], 2D[0];
>  1: MOV temp[1], input[3];
>  2: TEX temp[2], temp[1], CUBE[1];
>  3: MUL temp[2], temp[2], const[0];
>  4: MUL temp[3], temp[0], const[1];
>  5: MUL temp[3], temp[2], temp[3];
>  6: MUL temp[3], const[2].xxxx, temp[3];
>  7: ADD temp[4], const[3], -temp[3];
>  8: MAD temp[3].xyz, input[1].xxxx, temp[4], temp[3];
>  9: MUL output[0], temp[3], input[0];
> 10: MOV output[0].w, const[2].yyyy;
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: TEX temp[0], input[2], 2D[0];
>  1: MOV temp[1], input[3];
>  2: TEX temp[2], temp[1], CUBE[1];
>  3: MUL temp[2], temp[2], const[0];
>  4: MUL temp[3], temp[0], const[1];
>  5: MUL temp[3], temp[2], temp[3];
>  6: MUL temp[3], const[2].xxxx, temp[3];
>  7: ADD temp[4], const[3], -temp[3];
>  8: MAD temp[3].xyz, input[1].xxxx, temp[4], temp[3];
>  9: MUL output[0], temp[3], input[0];
> 10: MOV output[0].w, const[2].yyyy;
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: TEX temp[0].xyz, input[2].xy__, 2D[0];
>  1: MOV temp[1].xyz, input[3].xyz_;
>  2: TEX temp[2].xyz, temp[1].xyz_, CUBE[1];
>  3: MUL temp[2].xyz, temp[2].xyz_, const[0].xyz_;
>  4: MUL temp[3].xyz, temp[0].xyz_, const[1].xyz_;
>  5: MUL temp[3].xyz, temp[2].xyz_, temp[3].xyz_;
>  6: MUL temp[3].xyz, const[2].xxx_, temp[3].xyz_;
>  7: ADD temp[4].xyz, const[3].xyz_, -temp[3].xyz_;
>  8: MAD temp[3].xyz, input[1].xxx_, temp[4].xyz_, temp[3].xyz_;
>  9: MUL output[0].xyz, temp[3].xyz_, input[0].xyz_;
> 10: MOV output[0].w, const[2].___y;
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: TEX temp[0].xyz, input[2].xy__, 2D[0];
>  1: TEX temp[2].xyz, input[3].xyz_, CUBE[1];
>  2: MUL temp[2].xyz, temp[2].xyz_, const[0].xyz_;
>  3: MUL temp[3].xyz, temp[0].xyz_, const[1].xyz_;
>  4: MUL temp[3].xyz, temp[2].xyz_, temp[3].xyz_;
>  5: MUL temp[3].xyz, const[2].xxx_, temp[3].xyz_;
>  6: ADD temp[4].xyz, const[3].xyz_, -temp[3].xyz_;
>  7: MAD temp[3].xyz, input[1].xxx_, temp[4].xyz_, temp[3].xyz_;
>  8: MUL output[0].xyz, temp[3].xyz_, input[0].xyz_;
>  9: MOV output[0].w, const[2].___y;
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: TEX temp[0].xyz, input[2].xy__, 2D[0];
>  1: TEX temp[2].xyz, input[3].xyz_, CUBE[1];
>  2: MUL temp[2].xyz, temp[2].xyz_, const[0].xyz_;
>  3: MUL temp[3].xyz, temp[0].xyz_, const[1].xyz_;
>  4: MUL temp[3].xyz, temp[2].xyz_, temp[3].xyz_;
>  5: MUL temp[3].xyz, const[2].xxx_, temp[3].xyz_;
>  6: ADD temp[4].xyz, const[3].xyz_, -temp[3].xyz_;
>  7: MAD temp[3].xyz, input[1].xxx_, temp[4].xyz_, temp[3].xyz_;
>  8: MUL output[0].xyz, temp[3].xyz_, input[0].xyz_;
>  9: MOV output[0].w, const[2].___y;
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: TEX temp[0].xyz, input[2].xy__, 2D[0];
>  1: TEX temp[2].xyz, input[3].xyz_, CUBE[1];
>  2: src0.xyz = temp[2], src1.xyz = const[0]
>     MAD temp[2].xyz, src0.xyz, src1.xyz, src0.000
>  3: src0.xyz = temp[0], src1.xyz = const[1]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src0.000
>  4: src0.xyz = temp[2], src1.xyz = temp[3]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src0.000
>  5: src0.xyz = const[2], src1.xyz = temp[3]
>     MAD temp[3].xyz, src0.xxx, src1.xyz, src0.000
>  6: src0.xyz = const[3], src1.xyz = temp[3]
>     MAD temp[4].xyz, src0.xyz, src0.111, -src1.xyz
>  7: src0.xyz = input[1], src1.xyz = temp[4], src2.xyz = temp[3]
>     MAD temp[3].xyz, src0.xxx, src1.xyz, src2.xyz
>  8: src0.xyz = temp[3], src1.xyz = input[0]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>  9: src0.xyz = const[2]
>     MAD color[0].w, src0.y, src0.1, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[2].xyz, input[3].xyz_, CUBE[1];
>  2: TEX temp[0].xyz, input[2].xy__, 2D[0];
>  3: src0.xyz = temp[0], src1.xyz = const[1], src2.xyz = const[2]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src2.y, src0.1, src0.0
>  4: src0.xyz = temp[2], src1.xyz = const[0]
>     MAD temp[2].xyz, src0.xyz, src1.xyz, src0.000
>  5: src0.xyz = temp[2], src1.xyz = temp[3]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src0.000
>  6: src0.xyz = const[2], src1.xyz = temp[3]
>     MAD temp[3].xyz, src0.xxx, src1.xyz, src0.000
>  7: src0.xyz = const[3], src1.xyz = temp[3]
>     MAD temp[4].xyz, src0.xyz, src0.111, -src1.xyz
>  8: src0.xyz = input[1], src1.xyz = temp[4], src2.xyz = temp[3]
>     MAD temp[3].xyz, src0.xxx, src1.xyz, src2.xyz
>  9: src0.xyz = temp[3], src1.xyz = input[0]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[3].xyz, temp[3].xyz_, CUBE[1];
>  2: TEX temp[2].xyz, temp[2].xy__, 2D[0];
>  3: src0.xyz = temp[2], src1.xyz = const[1], src2.xyz = const[2]
>     MAD temp[2].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src2.y, src0.1, src0.0
>  4: src0.xyz = temp[3], src1.xyz = const[0]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src0.000
>  5: src0.xyz = temp[3], src1.xyz = temp[2]
>     MAD temp[2].xyz, src0.xyz, src1.xyz, src0.000
>  6: src0.xyz = const[2], src1.xyz = temp[2]
>     MAD temp[2].xyz, src0.xxx, src1.xyz, src0.000
>  7: src0.xyz = const[3], src1.xyz = temp[2]
>     MAD temp[3].xyz, src0.xyz, src0.111, -src1.xyz
>  8: src0.xyz = temp[1], src1.xyz = temp[3], src2.xyz = temp[2]
>     MAD temp[2].xyz, src0.xxx, src1.xyz, src2.xyz
>  9: src0.xyz = temp[2], src1.xyz = temp[0]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>pc=32*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 6, tex_end: 1  (code_addr: 00420180)
>  TEX:
>    TEX t3, t3, texture[1]   (000088c3)
>    TEX t2, t2, texture[0]   (00008082)
>  0: xyz:  t2  c1  c2 -> t2.xyz               (038a2842)
>       w:  t0  t0  t0 -> o0.w                 (01000000)
>     xyz:   t2.xyz   c1.xyz      0.0    op: 00050200
>       w:     c2.y      1.0      0.0    op: 00040887
>  1: xyz:  t3  c0  t0 -> t3.xyz               (038c0803)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t3.xyz   c0.xyz      0.0    op: 00050200
>       w:     t3.x     t3.x     t3.x    op: 00000000
>  2: xyz:  t3  t2  t0 -> t2.xyz               (03880083)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t3.xyz   t2.xyz      0.0    op: 00050200
>       w:     t3.x     t3.x     t3.x    op: 00000000
>  3: xyz:  c2  t2  t0 -> t2.xyz               (038800a2)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   c2.xxx   t2.xyz      0.0    op: 00050201
>       w:     c2.x     c2.x     c2.x    op: 00000000
>  4: xyz:  c3  t2  t0 -> t3.xyz               (038c00a3)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   c3.xyz      1.0  -t2.xyz    op: 00090a80
>       w:     c3.x     c3.x     c3.x    op: 00000000
>  5: xyz:  t1  t3  t2 -> t2.xyz               (038820c1)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xxx   t3.xyz   t2.xyz    op: 00020201
>       w:     t1.x     t1.x     t1.x    op: 00000000
>  6: xyz:  t2  t0  t0 -> o0.xyz               (1c000002)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz   t0.xyz      0.0    op: 00050200
>       w:     t2.x     t2.x     t2.x    op: 00000000
>r300: Initial fragment program
>FRAG
>DCL IN[0], COLOR, LINEAR
>DCL OUT[0], COLOR
>  0: MOV OUT[0], IN[0]
>  1: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: src0.xyz = input[0], src0.w = input[0]
>     MAD color[0].xyz, src0.xyz, src0.111, src0.000
>     MAD color[0].w, src0.w, src0.1, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: src0.xyz = input[0], src0.w = input[0]
>     MAD color[0].xyz, src0.xyz, src0.111, src0.000
>     MAD color[0].w, src0.w, src0.1, src0.0
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: src0.xyz = temp[0], src0.w = temp[0]
>     MAD color[0].xyz, src0.xyz, src0.111, src0.000
>     MAD color[0].w, src0.w, src0.1, src0.0
>pc=33*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 0, tex_end: 0  (code_addr: 00400000)
>  0: xyz:  t0  t0  t0 -> o0.xyz               (1c000000)
>       w:  t0  t0  t0 -> o0.w                 (01000000)
>     xyz:   t0.xyz      1.0      0.0    op: 00050a80
>       w:     t0.w      1.0      0.0    op: 00040889
>r300: Initial fragment program
>FRAG
>DCL IN[0], COLOR, LINEAR
>DCL IN[1], GENERIC[0], PERSPECTIVE
>DCL IN[2], GENERIC[1], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL SAMP[1]
>DCL TEMP[0..1]
>  0: TEX TEMP[0], IN[1], SAMP[0], 2D
>  1: TEX TEMP[1], IN[2], SAMP[1], 2D
>  2: MUL TEMP[0], IN[0], TEMP[0]
>  3: MUL OUT[0], TEMP[1], TEMP[0]
>  4: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: TEX temp[1], input[2], 2D[1];
>  2: MUL temp[0], input[0], temp[0];
>  3: MUL output[0], temp[1], temp[0];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: TEX temp[1], input[2], 2D[1];
>  2: MUL temp[0], input[0], temp[0];
>  3: MUL output[0], temp[1], temp[0];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: TEX temp[1], input[2], 2D[1];
>  2: MUL temp[0], input[0], temp[0];
>  3: MUL output[0], temp[1], temp[0];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: TEX temp[1], input[2], 2D[1];
>  2: MUL temp[0], input[0], temp[0];
>  3: MUL output[0], temp[1], temp[0];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: TEX temp[1], input[2], 2D[1];
>  2: MUL temp[0], input[0], temp[0];
>  3: MUL output[0], temp[1], temp[0];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: TEX temp[0], input[1].xy__, 2D[0];
>  1: TEX temp[1], input[2].xy__, 2D[1];
>  2: MUL temp[0], input[0], temp[0];
>  3: MUL output[0], temp[1], temp[0];
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: TEX temp[0], input[1].xy__, 2D[0];
>  1: TEX temp[1], input[2].xy__, 2D[1];
>  2: MUL temp[0], input[0], temp[0];
>  3: MUL output[0], temp[1], temp[0];
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: TEX temp[0], input[1].xy__, 2D[0];
>  1: TEX temp[1], input[2].xy__, 2D[1];
>  2: MUL temp[0], input[0], temp[0];
>  3: MUL output[0], temp[1], temp[0];
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: TEX temp[0], input[1].xy__, 2D[0];
>  1: TEX temp[1], input[2].xy__, 2D[1];
>  2: src0.xyz = input[0], src0.w = input[0], src1.xyz = temp[0], src1.w = temp[0]
>     MAD temp[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD temp[0].w, src0.w, src1.w, src0.0
>  3: src0.xyz = temp[1], src0.w = temp[1], src1.xyz = temp[0], src1.w = temp[0]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[1], input[2].xy__, 2D[1];
>  2: TEX temp[0], input[1].xy__, 2D[0];
>  3: src0.xyz = input[0], src0.w = input[0], src1.xyz = temp[0], src1.w = temp[0]
>     MAD temp[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD temp[0].w, src0.w, src1.w, src0.0
>  4: src0.xyz = temp[1], src0.w = temp[1], src1.xyz = temp[0], src1.w = temp[0]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[2], temp[2].xy__, 2D[1];
>  2: TEX temp[1], temp[1].xy__, 2D[0];
>  3: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = temp[1], src1.w = temp[1]
>     MAD temp[1].xyz, src0.xyz, src1.xyz, src0.000
>     MAD temp[1].w, src0.w, src1.w, src0.0
>  4: src0.xyz = temp[2], src0.w = temp[2], src1.xyz = temp[1], src1.w = temp[1]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>pc=34*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 1, tex_end: 1  (code_addr: 00420040)
>  TEX:
>    TEX t2, t2, texture[1]   (00008882)
>    TEX t1, t1, texture[0]   (00008041)
>  0: xyz:  t0  t1  t0 -> t1.xyz               (03840040)
>       w:  t0  t1  t0 -> t1.w                 (00840040)
>     xyz:   t0.xyz   t1.xyz      0.0    op: 00050200
>       w:     t0.w     t1.w      0.0    op: 00040509
>  1: xyz:  t2  t1  t0 -> o0.xyz               (1c000042)
>       w:  t2  t1  t0 -> o0.w                 (01000042)
>     xyz:   t2.xyz   t1.xyz      0.0    op: 00050200
>       w:     t2.w     t1.w      0.0    op: 00040509
>r300: Initial fragment program
>FRAG
>DCL IN[0], COLOR, LINEAR
>DCL IN[1], GENERIC[0], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL SAMP[1]
>DCL CONST[1]
>DCL TEMP[0..1]
>IMM FLT32 {    0.7500,     1.0000,     0.0000,     0.0000}
>  0: ADD TEMP[0].xy, IN[1], IMM[0].xyzy
>  1: TEX TEMP[1], IN[1], SAMP[1], 2D
>  2: TEX TEMP[0], TEMP[0], SAMP[0], 2D
>  3: ADD TEMP[0], TEMP[0], -TEMP[1]
>  4: MAD TEMP[0], IN[0].xxxx, TEMP[0], TEMP[1]
>  5: MUL OUT[0], TEMP[0], CONST[1]
>  6: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: ADD temp[0].xy, input[1], const[2].xyzy;
>  1: TEX temp[1], input[1], 2D[1];
>  2: TEX temp[0], temp[0], 2D[0];
>  3: ADD temp[0], temp[0], -temp[1];
>  4: MAD temp[0], input[0].xxxx, temp[0], temp[1];
>  5: MUL output[0], temp[0], const[1];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: ADD temp[0].xy, input[1], const[2].xyzy;
>  1: TEX temp[1], input[1], 2D[1];
>  2: TEX temp[0], temp[0], 2D[0];
>  3: ADD temp[0], temp[0], -temp[1];
>  4: MAD temp[0], input[0].xxxx, temp[0], temp[1];
>  5: MUL output[0], temp[0], const[1];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: ADD temp[0].xy, input[1], const[2].xyzy;
>  1: TEX temp[1], input[1], 2D[1];
>  2: TEX temp[0], temp[0], 2D[0];
>  3: ADD temp[0], temp[0], -temp[1];
>  4: MAD temp[0], input[0].xxxx, temp[0], temp[1];
>  5: MUL output[0], temp[0], const[1];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: ADD temp[0].xy, input[1], const[2].xyzy;
>  1: TEX temp[1], input[1], 2D[1];
>  2: TEX temp[0], temp[0], 2D[0];
>  3: ADD temp[0], temp[0], -temp[1];
>  4: MAD temp[0], input[0].xxxx, temp[0], temp[1];
>  5: MUL output[0], temp[0], const[1];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: ADD temp[0].xy, input[1], const[2].xyzy;
>  1: TEX temp[1], input[1], 2D[1];
>  2: TEX temp[0], temp[0], 2D[0];
>  3: ADD temp[0], temp[0], -temp[1];
>  4: MAD temp[0], input[0].xxxx, temp[0], temp[1];
>  5: MUL output[0], temp[0], const[1];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: ADD temp[0].xy, input[1].xy__, const[2].xy__;
>  1: TEX temp[1], input[1].xy__, 2D[1];
>  2: TEX temp[0], temp[0].xy__, 2D[0];
>  3: ADD temp[0], temp[0], -temp[1];
>  4: MAD temp[0], input[0].xxxx, temp[0], temp[1];
>  5: MUL output[0], temp[0], const[1];
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: ADD temp[0].xy, input[1].xy__, const[2].xy__;
>  1: TEX temp[1], input[1].xy__, 2D[1];
>  2: TEX temp[0], temp[0].xy__, 2D[0];
>  3: ADD temp[0], temp[0], -temp[1];
>  4: MAD temp[0], input[0].xxxx, temp[0], temp[1];
>  5: MUL output[0], temp[0], const[1];
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: ADD temp[0].xy, input[1].xy__, const[2].xy__;
>  1: TEX temp[1], input[1].xy__, 2D[1];
>  2: TEX temp[0], temp[0].xy__, 2D[0];
>  3: ADD temp[0], temp[0], -temp[1];
>  4: MAD temp[0], input[0].xxxx, temp[0], temp[1];
>  5: MUL output[0], temp[0], const[1];
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: src0.xyz = input[1], src1.xyz = const[2]
>     MAD temp[0].xy, src0.xy_, src0.111, src1.xy_
>  1: TEX temp[1], input[1].xy__, 2D[1];
>  2: TEX temp[0], temp[0].xy__, 2D[0];
>  3: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = temp[1], src1.w = temp[1]
>     MAD temp[0].xyz, src0.xyz, src0.111, -src1.xyz
>     MAD temp[0].w, src0.w, src0.1, -src1.w
>  4: src0.xyz = input[0], src0.w = temp[0], src1.xyz = temp[0], src1.w = temp[1], src2.xyz = temp[1]
>     MAD temp[0].xyz, src0.xxx, src1.xyz, src2.xyz
>     MAD temp[0].w, src0.x, src0.w, src1.w
>  5: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[1], src1.w = const[1]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[1], input[1].xy__, 2D[1];
>  2: src0.xyz = input[1], src1.xyz = const[2]
>     MAD temp[0].xy, src0.xy_, src0.111, src1.xy_
>  3: BEGIN_TEX;
>  4: TEX temp[0], temp[0].xy__, 2D[0];
>  5: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = temp[1], src1.w = temp[1]
>     MAD temp[0].xyz, src0.xyz, src0.111, -src1.xyz
>     MAD temp[0].w, src0.w, src0.1, -src1.w
>  6: src0.xyz = input[0], src0.w = temp[0], src1.xyz = temp[0], src1.w = temp[1], src2.xyz = temp[1]
>     MAD temp[0].xyz, src0.xxx, src1.xyz, src2.xyz
>     MAD temp[0].w, src0.x, src0.w, src1.w
>  7: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[1], src1.w = const[1]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[2], temp[1].xy__, 2D[1];
>  2: src0.xyz = temp[1], src1.xyz = const[2]
>     MAD temp[1].xy, src0.xy_, src0.111, src1.xy_
>  3: BEGIN_TEX;
>  4: TEX temp[1], temp[1].xy__, 2D[0];
>  5: src0.xyz = temp[1], src0.w = temp[1], src1.xyz = temp[2], src1.w = temp[2]
>     MAD temp[1].xyz, src0.xyz, src0.111, -src1.xyz
>     MAD temp[1].w, src0.w, src0.1, -src1.w
>  6: src0.xyz = temp[0], src0.w = temp[1], src1.xyz = temp[1], src1.w = temp[2], src2.xyz = temp[2]
>     MAD temp[1].xyz, src0.xxx, src1.xyz, src2.xyz
>     MAD temp[1].w, src0.x, src0.w, src1.w
>  7: src0.xyz = temp[1], src0.w = temp[1], src1.xyz = const[1], src1.w = const[1]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>pc=35*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 0, tex_end: 0  (code_addr: 00000000)
>  TEX:
>    TEX t2, t1, texture[1]   (00008881)
>  0: xyz:  t1  c2  t0 -> t1.xy                (01840881)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz      1.0   c2.xyz    op: 00010a80
>       w:     t1.x     t1.x     t1.x    op: 00000000
>NODE 1: alu_offset: 1, tex_offset: 1, alu_end: 2, tex_end: 0  (code_addr: 00401081)
>  TEX:
>    TEX t1, t1, texture[0]   (00008041)
>  1: xyz:  t1  t2  t0 -> t1.xyz               (03840081)
>       w:  t1  t2  t0 -> t1.w                 (00840081)
>     xyz:   t1.xyz      1.0  -t2.xyz    op: 00090a80
>       w:     t1.w      1.0    -t2.w    op: 000a8889
>  2: xyz:  t0  t1  t2 -> t1.xyz               (03842040)
>       w:  t1  t2  t0 -> t1.w                 (00840081)
>     xyz:   t0.xxx   t1.xyz   t2.xyz    op: 00020201
>       w:     t0.x     t1.w     t2.w    op: 00028480
>  3: xyz:  t1  c1  t0 -> o0.xyz               (1c000841)
>       w:  t1  c1  t0 -> o0.w                 (01000841)
>     xyz:   t1.xyz   c1.xyz      0.0    op: 00050200
>       w:     t1.w     c1.w      0.0    op: 00040509
>r300: Initial fragment program
>FRAG
>DCL IN[0], COLOR, LINEAR
>DCL IN[1], COLOR[1], LINEAR
>DCL OUT[0], COLOR
>DCL CONST[0]
>  0: MOV OUT[0], IN[0]
>  1: LRP OUT[0].xyz, IN[1].xxxx, CONST[0], IN[0]
>  2: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>  1: LRP output[0].xyz, input[1].xxxx, const[0], input[0];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>  1: LRP output[0].xyz, input[1].xxxx, const[0], input[0];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>  1: LRP output[0].xyz, input[1].xxxx, const[0], input[0];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>  1: ADD temp[0], const[0], -input[0];
>  2: MAD output[0].xyz, input[1].xxxx, temp[0], input[0];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: MOV output[0], input[0];
>  1: ADD temp[0], const[0], -input[0];
>  2: MAD output[0].xyz, input[1].xxxx, temp[0], input[0];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: MOV output[0].w, input[0].___w;
>  1: ADD temp[0].xyz, const[0].xyz_, -input[0].xyz_;
>  2: MAD output[0].xyz, input[1].xxx_, temp[0].xyz_, input[0].xyz_;
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: MOV output[0].w, input[0].___w;
>  1: ADD temp[0].xyz, const[0].xyz_, -input[0].xyz_;
>  2: MAD output[0].xyz, input[1].xxx_, temp[0].xyz_, input[0].xyz_;
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: MOV output[0].w, input[0].___w;
>  1: ADD temp[0].xyz, const[0].xyz_, -input[0].xyz_;
>  2: MAD output[0].xyz, input[1].xxx_, temp[0].xyz_, input[0].xyz_;
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: src0.w = input[0]
>     MAD color[0].w, src0.w, src0.1, src0.0
>  1: src0.xyz = const[0], src1.xyz = input[0]
>     MAD temp[0].xyz, src0.xyz, src0.111, -src1.xyz
>  2: src0.xyz = input[1], src1.xyz = temp[0], src2.xyz = input[0]
>     MAD color[0].xyz, src0.xxx, src1.xyz, src2.xyz
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: src0.xyz = const[0], src0.w = input[0], src1.xyz = input[0]
>     MAD temp[0].xyz, src0.xyz, src0.111, -src1.xyz
>     MAD color[0].w, src0.w, src0.1, src0.0
>  1: src0.xyz = input[1], src1.xyz = temp[0], src2.xyz = input[0]
>     MAD color[0].xyz, src0.xxx, src1.xyz, src2.xyz
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: src0.xyz = const[0], src0.w = temp[0], src1.xyz = temp[0]
>     MAD temp[2].xyz, src0.xyz, src0.111, -src1.xyz
>     MAD color[0].w, src0.w, src0.1, src0.0
>  1: src0.xyz = temp[1], src1.xyz = temp[2], src2.xyz = temp[0]
>     MAD color[0].xyz, src0.xxx, src1.xyz, src2.xyz
>pc=36*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 1, tex_end: 0  (code_addr: 00400040)
>  0: xyz:  c0  t0  t0 -> t2.xyz               (03880020)
>       w:  t0  t0  t0 -> o0.w                 (01000000)
>     xyz:   c0.xyz      1.0  -t0.xyz    op: 00090a80
>       w:     t0.w      1.0      0.0    op: 00040889
>  1: xyz:  t1  t2  t0 -> o0.xyz               (1c000081)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xxx   t2.xyz   t0.xyz    op: 00020201
>       w:     t1.x     t1.x     t1.x    op: 00000000
>r300: Initial fragment program
>FRAG
>DCL IN[0], COLOR, LINEAR
>DCL IN[1], GENERIC[0], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL CONST[1]
>DCL TEMP[0]
>IMM FLT32 {    1.0000,     0.0000,     0.0000,     0.0000}
>  0: TEX TEMP[0], IN[1], SAMP[0], 2D
>  1: SUB TEMP[0], TEMP[0], IMM[0].xxxx
>  2: MUL TEMP[0], TEMP[0], IN[0]
>  3: MAD OUT[0].xyz, TEMP[0], CONST[1], IMM[0].xxxx
>  4: MOV OUT[0].w, IMM[0].xxxx
>  5: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: SUB temp[0], temp[0], temp[0].1111;
>  2: MUL temp[0], temp[0], input[0];
>  3: MAD output[0].xyz, temp[0], const[1], temp[0].1111;
>  4: MOV output[0].w, temp[0].1111;
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: SUB temp[0], temp[0], temp[0].1111;
>  2: MUL temp[0], temp[0], input[0];
>  3: MAD output[0].xyz, temp[0], const[1], temp[0].1111;
>  4: MOV output[0].w, temp[0].1111;
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: SUB temp[0], temp[0], temp[0].1111;
>  2: MUL temp[0], temp[0], input[0];
>  3: MAD output[0].xyz, temp[0], const[1], temp[0].1111;
>  4: MOV output[0].w, temp[0].1111;
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: ADD temp[0], temp[0], -temp[0].1111;
>  2: MUL temp[0], temp[0], input[0];
>  3: MAD output[0].xyz, temp[0], const[1], temp[0].1111;
>  4: MOV output[0].w, temp[0].1111;
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: ADD temp[0], temp[0], -temp[0].1111;
>  2: MUL temp[0], temp[0], input[0];
>  3: MAD output[0].xyz, temp[0], const[1], temp[0].1111;
>  4: MOV output[0].w, temp[0].1111;
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: TEX temp[0].xyz, input[1].xy__, 2D[0];
>  1: ADD temp[0].xyz, temp[0].xyz_, -temp[0].111_;
>  2: MUL temp[0].xyz, temp[0].xyz_, input[0].xyz_;
>  3: MAD output[0].xyz, temp[0].xyz_, const[1].xyz_, temp[0].111_;
>  4: MOV output[0].w, temp[0].___1;
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: TEX temp[0].xyz, input[1].xy__, 2D[0];
>  1: ADD temp[0].xyz, temp[0].xyz_, -temp[0].111_;
>  2: MUL temp[0].xyz, temp[0].xyz_, input[0].xyz_;
>  3: MAD output[0].xyz, temp[0].xyz_, const[1].xyz_, temp[0].111_;
>  4: MOV output[0].w, temp[0].___1;
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: TEX temp[0].xyz, input[1].xy__, 2D[0];
>  1: ADD temp[0].xyz, temp[0].xyz_, -temp[0].111_;
>  2: MUL temp[0].xyz, temp[0].xyz_, input[0].xyz_;
>  3: MAD output[0].xyz, temp[0].xyz_, const[1].xyz_, temp[0].111_;
>  4: MOV output[0].w, temp[0].___1;
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: TEX temp[0].xyz, input[1].xy__, 2D[0];
>  1: src0.xyz = temp[0]
>     MAD temp[0].xyz, src0.xyz, src0.111, -src0.111
>  2: src0.xyz = temp[0], src1.xyz = input[0]
>     MAD temp[0].xyz, src0.xyz, src1.xyz, src0.000
>  3: src0.xyz = temp[0], src1.xyz = const[1]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.111
>  4: 
>     MAD color[0].w, src0.1, src0.1, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[0].xyz, input[1].xy__, 2D[0];
>  2: src0.xyz = temp[0]
>     MAD temp[0].xyz, src0.xyz, src0.111, -src0.111
>     MAD color[0].w, src0.1, src0.1, src0.0
>  3: src0.xyz = temp[0], src1.xyz = input[0]
>     MAD temp[0].xyz, src0.xyz, src1.xyz, src0.000
>  4: src0.xyz = temp[0], src1.xyz = const[1]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.111
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[1].xyz, temp[1].xy__, 2D[0];
>  2: src0.xyz = temp[1]
>     MAD temp[1].xyz, src0.xyz, src0.111, -src0.111
>     MAD color[0].w, src0.1, src0.1, src0.0
>  3: src0.xyz = temp[1], src1.xyz = temp[0]
>     MAD temp[1].xyz, src0.xyz, src1.xyz, src0.000
>  4: src0.xyz = temp[1], src1.xyz = const[1]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.111
>pc=37*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 2, tex_end: 0  (code_addr: 00400080)
>  TEX:
>    TEX t1, t1, texture[0]   (00008041)
>  0: xyz:  t1  t0  t0 -> t1.xyz               (03840001)
>       w:  t0  t0  t0 -> o0.w                 (01000000)
>     xyz:   t1.xyz      1.0     -1.0    op: 000d4a80
>       w:      1.0      1.0      0.0    op: 00040891
>  1: xyz:  t1  t0  t0 -> t1.xyz               (03840001)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   t0.xyz      0.0    op: 00050200
>       w:     t1.x     t1.x     t1.x    op: 00000000
>  2: xyz:  t1  c1  t0 -> o0.xyz               (1c000841)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   c1.xyz      1.0    op: 00054200
>       w:     t1.x     t1.x     t1.x    op: 00000000
>r300: Initial fragment program
>FRAG
>DCL IN[0], POSITION, LINEAR
>DCL IN[1], COLOR, LINEAR
>DCL IN[2], COLOR[1], LINEAR
>DCL IN[3], GENERIC[0], PERSPECTIVE
>DCL IN[4], GENERIC[1], PERSPECTIVE
>DCL IN[5], GENERIC[4], PERSPECTIVE
>DCL IN[6], GENERIC[5], PERSPECTIVE
>DCL IN[7], GENERIC[6], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL SAMP[1]
>DCL SAMP[2]
>DCL CONST[4..5]
>DCL CONST[1..3]
>DCL TEMP[0..2]
>IMM FLT32 {    2.0000,     1.0000,     0.5000,     0.0625}
>  0: MOV TEMP[0], IN[0]
>  1: SUB TEMP[0].y, CONST[5], IN[0]
>  2: TEX TEMP[1], IN[4], SAMP[1], 2D
>  3: MUL TEMP[1].xyz, TEMP[1], IMM[0].xxxx
>  4: ADD TEMP[1].xyz, TEMP[1], -IMM[0].yyyy
>  5: MUL TEMP[2].xyz, IN[6], TEMP[1].yyyy
>  6: MAD TEMP[2].xyz, IN[5], -TEMP[1].xxxx, TEMP[2]
>  7: MAD TEMP[1].xyz, IN[7], TEMP[1].zzzz, TEMP[2]
>  8: DP3 TEMP[2].x, TEMP[1], TEMP[1]
>  9: RSQ TEMP[2].x, TEMP[2].xxxx
> 10: MUL TEMP[2].xyz, TEMP[2].xxxx, TEMP[1]
> 11: TEX TEMP[1].xyz, TEMP[2], SAMP[2], CUBE
> 12: DP3 TEMP[2].x, TEMP[2], CONST[1]
> 13: MAX TEMP[1].w, TEMP[2].xxxx, TEMP[1]
> 14: MUL TEMP[1].xyz, TEMP[1], CONST[2].xxxx
> 15: MAD TEMP[2].xyz, TEMP[1].wwww, CONST[3], TEMP[1]
> 16: TEX TEMP[1], IN[3], SAMP[1], 2D
> 17: MUL TEMP[1].xyz, TEMP[1], TEMP[2]
> 18: ADD TEMP[2].xyz, -TEMP[1], CONST[4]
> 19: MUL TEMP[2].xyz, IN[2].xxxx, TEMP[2]
> 20: ADD OUT[0].xyz, TEMP[1], TEMP[2]
> 21: MAD TEMP[1].xy, TEMP[0], IMM[0].wwww, IMM[0].wwww
> 22: TEX TEMP[2].w, TEMP[1], SAMP[0], 2D
> 23: SLT TEMP[1].x, TEMP[1].wwww, IMM[0].zzzz
> 24: SLT TEMP[1].y, IN[1].wwww, TEMP[2].wwww
> 25: MOV OUT[0].w, TEMP[1]
> 26: KIL -TEMP[1].xxxx
> 27: KIL -TEMP[1].yyyy
> 28: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: RCP temp[3].w, input[0].wwww;
>  1: MUL temp[3].xyz, input[0], temp[3].wwww;
>  2: MAD temp[3].xyz, temp[3].xyz0, const[7].xyz0, const[8].xyz0;
>  3: MOV temp[0], temp[3];
>  4: SUB temp[0].y, const[5], temp[3];
>  5: TEX temp[1], input[4], 2D[1];
>  6: MUL temp[1].xyz, temp[1], const[6].xxxx;
>  7: ADD temp[1].xyz, temp[1], -const[6].yyyy;
>  8: MUL temp[2].xyz, input[6], temp[1].yyyy;
>  9: MAD temp[2].xyz, input[5], -temp[1].xxxx, temp[2];
> 10: MAD temp[1].xyz, input[7], temp[1].zzzz, temp[2];
> 11: DP3 temp[2].x, temp[1], temp[1];
> 12: RSQ temp[2].x, temp[2].xxxx;
> 13: MUL temp[2].xyz, temp[2].xxxx, temp[1];
> 14: TEX temp[1].xyz, temp[2], CUBE[2];
> 15: DP3 temp[2].x, temp[2], const[1];
> 16: MAX temp[1].w, temp[2].xxxx, temp[1];
> 17: MUL temp[1].xyz, temp[1], const[2].xxxx;
> 18: MAD temp[2].xyz, temp[1].wwww, const[3], temp[1];
> 19: TEX temp[1], input[3], 2D[1];
> 20: MUL temp[1].xyz, temp[1], temp[2];
> 21: ADD temp[2].xyz, -temp[1], const[4];
> 22: MUL temp[2].xyz, input[2].xxxx, temp[2];
> 23: ADD output[0].xyz, temp[1], temp[2];
> 24: MAD temp[1].xy, temp[0], const[6].wwww, const[6].wwww;
> 25: TEX temp[2].w, temp[1], 2D[0];
> 26: SLT temp[1].x, temp[1].wwww, const[6].zzzz;
> 27: SLT temp[1].y, input[1].wwww, temp[2].wwww;
> 28: MOV output[0].w, temp[1];
> 29: KIL -temp[1].xxxx;
> 30: KIL -temp[1].yyyy;
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: RCP temp[3].w, input[0].wwww;
>  1: MUL temp[3].xyz, input[0], temp[3].wwww;
>  2: MAD temp[3].xyz, temp[3].xyz0, const[7].xyz0, const[8].xyz0;
>  3: MOV temp[0], temp[3];
>  4: SUB temp[0].y, const[5], temp[3];
>  5: TEX temp[1], input[4], 2D[1];
>  6: MUL temp[1].xyz, temp[1], const[6].xxxx;
>  7: ADD temp[1].xyz, temp[1], -const[6].yyyy;
>  8: MUL temp[2].xyz, input[6], temp[1].yyyy;
>  9: MAD temp[2].xyz, input[5], -temp[1].xxxx, temp[2];
> 10: MAD temp[1].xyz, input[7], temp[1].zzzz, temp[2];
> 11: DP3 temp[2].x, temp[1], temp[1];
> 12: RSQ temp[2].x, temp[2].xxxx;
> 13: MUL temp[2].xyz, temp[2].xxxx, temp[1];
> 14: TEX temp[1].xyz, temp[2], CUBE[2];
> 15: DP3 temp[2].x, temp[2], const[1];
> 16: MAX temp[1].w, temp[2].xxxx, temp[1];
> 17: MUL temp[1].xyz, temp[1], const[2].xxxx;
> 18: MAD temp[2].xyz, temp[1].wwww, const[3], temp[1];
> 19: TEX temp[1], input[3], 2D[1];
> 20: MUL temp[1].xyz, temp[1], temp[2];
> 21: ADD temp[2].xyz, -temp[1], const[4];
> 22: MUL temp[2].xyz, input[2].xxxx, temp[2];
> 23: ADD output[0].xyz, temp[1], temp[2];
> 24: MAD temp[1].xy, temp[0], const[6].wwww, const[6].wwww;
> 25: TEX temp[2].w, temp[1], 2D[0];
> 26: SLT temp[1].x, temp[1].wwww, const[6].zzzz;
> 27: SLT temp[1].y, input[1].wwww, temp[2].wwww;
> 28: MOV output[0].w, temp[1];
> 29: KIL -temp[1].xxxx;
> 30: KIL -temp[1].yyyy;
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: RCP temp[3].w, input[0].wwww;
>  1: MUL temp[3].xyz, input[0], temp[3].wwww;
>  2: MAD temp[3].xyz, temp[3].xyz0, const[7].xyz0, const[8].xyz0;
>  3: MOV temp[0], temp[3];
>  4: SUB temp[0].y, const[5], temp[3];
>  5: TEX temp[1], input[4], 2D[1];
>  6: MUL temp[1].xyz, temp[1], const[6].xxxx;
>  7: ADD temp[1].xyz, temp[1], -const[6].yyyy;
>  8: MUL temp[2].xyz, input[6], temp[1].yyyy;
>  9: MAD temp[2].xyz, input[5], -temp[1].xxxx, temp[2];
> 10: MAD temp[1].xyz, input[7], temp[1].zzzz, temp[2];
> 11: DP3 temp[2].x, temp[1], temp[1];
> 12: RSQ temp[2].x, temp[2].xxxx;
> 13: MUL temp[2].xyz, temp[2].xxxx, temp[1];
> 14: TEX temp[1].xyz, temp[2], CUBE[2];
> 15: DP3 temp[2].x, temp[2], const[1];
> 16: MAX temp[1].w, temp[2].xxxx, temp[1];
> 17: MUL temp[1].xyz, temp[1], const[2].xxxx;
> 18: MAD temp[2].xyz, temp[1].wwww, const[3], temp[1];
> 19: TEX temp[1], input[3], 2D[1];
> 20: MUL temp[1].xyz, temp[1], temp[2];
> 21: ADD temp[2].xyz, -temp[1], const[4];
> 22: MUL temp[2].xyz, input[2].xxxx, temp[2];
> 23: ADD output[0].xyz, temp[1], temp[2];
> 24: MAD temp[1].xy, temp[0], const[6].wwww, const[6].wwww;
> 25: TEX temp[2].w, temp[1], 2D[0];
> 26: SLT temp[1].x, temp[1].wwww, const[6].zzzz;
> 27: SLT temp[1].y, input[1].wwww, temp[2].wwww;
> 28: MOV output[0].w, temp[1];
> 29: KIL -temp[1].xxxx;
> 30: KIL -temp[1].yyyy;
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: RCP temp[3].w, input[0].wwww;
>  1: MUL temp[3].xyz, input[0], temp[3].wwww;
>  2: MAD temp[3].xyz, temp[3].xyz0, const[7].xyz0, const[8].xyz0;
>  3: MOV temp[0], temp[3];
>  4: ADD temp[0].y, const[5], -temp[3];
>  5: TEX temp[1], input[4], 2D[1];
>  6: MUL temp[1].xyz, temp[1], const[6].xxxx;
>  7: ADD temp[1].xyz, temp[1], -const[6].yyyy;
>  8: MUL temp[2].xyz, input[6], temp[1].yyyy;
>  9: MAD temp[2].xyz, input[5], -temp[1].xxxx, temp[2];
> 10: MAD temp[1].xyz, input[7], temp[1].zzzz, temp[2];
> 11: DP3 temp[2].x, temp[1], temp[1];
> 12: RSQ temp[2].x, |temp[2].xxxx|;
> 13: MUL temp[2].xyz, temp[2].xxxx, temp[1];
> 14: TEX temp[1].xyz, temp[2], CUBE[2];
> 15: DP3 temp[2].x, temp[2], const[1];
> 16: MAX temp[1].w, temp[2].xxxx, temp[1];
> 17: MUL temp[1].xyz, temp[1], const[2].xxxx;
> 18: MAD temp[2].xyz, temp[1].wwww, const[3], temp[1];
> 19: TEX temp[1], input[3], 2D[1];
> 20: MUL temp[1].xyz, temp[1], temp[2];
> 21: ADD temp[2].xyz, -temp[1], const[4];
> 22: MUL temp[2].xyz, input[2].xxxx, temp[2];
> 23: ADD output[0].xyz, temp[1], temp[2];
> 24: MAD temp[1].xy, temp[0], const[6].wwww, const[6].wwww;
> 25: TEX temp[2].w, temp[1], 2D[0];
> 26: ADD temp[4], temp[1].wwww, -const[6].zzzz;
> 27: CMP temp[1].x, temp[4], none.1111, none.0000;
> 28: ADD temp[5], input[1].wwww, -temp[2].wwww;
> 29: CMP temp[1].y, temp[5], none.1111, none.0000;
> 30: MOV output[0].w, temp[1];
> 31: KIL -temp[1].xxxx;
> 32: KIL -temp[1].yyyy;
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: RCP temp[3].w, input[0].wwww;
>  1: MUL temp[3].xyz, input[0], temp[3].wwww;
>  2: MAD temp[3].xyz, temp[3].xyz0, const[7].xyz0, const[8].xyz0;
>  3: MOV temp[0], temp[3];
>  4: ADD temp[0].y, const[5], -temp[3];
>  5: TEX temp[1], input[4], 2D[1];
>  6: MUL temp[1].xyz, temp[1], const[6].xxxx;
>  7: ADD temp[1].xyz, temp[1], -const[6].yyyy;
>  8: MUL temp[2].xyz, input[6], temp[1].yyyy;
>  9: MAD temp[2].xyz, input[5], -temp[1].xxxx, temp[2];
> 10: MAD temp[1].xyz, input[7], temp[1].zzzz, temp[2];
> 11: DP3 temp[2].x, temp[1], temp[1];
> 12: RSQ temp[2].x, |temp[2].xxxx|;
> 13: MUL temp[2].xyz, temp[2].xxxx, temp[1];
> 14: TEX temp[6], temp[2], CUBE[2];
> 15: MOV temp[1].xyz, temp[6];
> 16: DP3 temp[2].x, temp[2], const[1];
> 17: MAX temp[1].w, temp[2].xxxx, temp[1];
> 18: MUL temp[1].xyz, temp[1], const[2].xxxx;
> 19: MAD temp[2].xyz, temp[1].wwww, const[3], temp[1];
> 20: TEX temp[1], input[3], 2D[1];
> 21: MUL temp[1].xyz, temp[1], temp[2];
> 22: ADD temp[2].xyz, -temp[1], const[4];
> 23: MUL temp[2].xyz, input[2].xxxx, temp[2];
> 24: ADD output[0].xyz, temp[1], temp[2];
> 25: MAD temp[1].xy, temp[0], const[6].wwww, const[6].wwww;
> 26: TEX temp[7], temp[1], 2D[0];
> 27: MOV temp[2].w, temp[7];
> 28: ADD temp[4], temp[1].wwww, -const[6].zzzz;
> 29: CMP temp[1].x, temp[4], none.1111, none.0000;
> 30: ADD temp[5], input[1].wwww, -temp[2].wwww;
> 31: CMP temp[1].y, temp[5], none.1111, none.0000;
> 32: MOV output[0].w, temp[1];
> 33: KIL -temp[1].xxxx;
> 34: KIL -temp[1].yyyy;
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: RCP temp[3].w, input[0].w___;
>  1: MUL temp[3].xy, input[0].xy__, temp[3].ww__;
>  2: MAD temp[3].xy, temp[3].xy__, const[7].xy__, const[8].xy__;
>  3: MOV temp[0].x, temp[3].x___;
>  4: ADD temp[0].y, const[5]._y__, -temp[3]._y__;
>  5: TEX temp[1], input[4].xy__, 2D[1];
>  6: MUL temp[1].xyz, temp[1].xyz_, const[6].xxx_;
>  7: ADD temp[1].xyz, temp[1].xyz_, -const[6].yyy_;
>  8: MUL temp[2].xyz, input[6].xyz_, temp[1].yyy_;
>  9: MAD temp[2].xyz, input[5].xyz_, -temp[1].xxx_, temp[2].xyz_;
> 10: MAD temp[1].xyz, input[7].xyz_, temp[1].zzz_, temp[2].xyz_;
> 11: DP3 temp[2].x, temp[1].xyz_, temp[1].xyz_;
> 12: RSQ temp[2].x, |temp[2].x___|;
> 13: MUL temp[2].xyz, temp[2].xxx_, temp[1].xyz_;
> 14: TEX temp[6].xyz, temp[2].xyz_, CUBE[2];
> 15: MOV temp[1].xyz, temp[6].xyz_;
> 16: DP3 temp[2].x, temp[2].xyz_, const[1].xyz_;
> 17: MAX temp[1].w, temp[2].___x, temp[1].___w;
> 18: MUL temp[1].xyz, temp[1].xyz_, const[2].xxx_;
> 19: MAD temp[2].xyz, temp[1].www_, const[3].xyz_, temp[1].xyz_;
> 20: TEX temp[1], input[3].xy__, 2D[1];
> 21: MUL temp[1].xyz, temp[1].xyz_, temp[2].xyz_;
> 22: ADD temp[2].xyz, -temp[1].xyz_, const[4].xyz_;
> 23: MUL temp[2].xyz, input[2].xxx_, temp[2].xyz_;
> 24: ADD output[0].xyz, temp[1].xyz_, temp[2].xyz_;
> 25: MAD temp[1].xy, temp[0].xy__, const[6].ww__, const[6].ww__;
> 26: TEX temp[7].w, temp[1].xy__, 2D[0];
> 27: MOV temp[2].w, temp[7].___w;
> 28: ADD temp[4].x, temp[1].w___, -const[6].z___;
> 29: CMP temp[1].x, temp[4].x___, none.1___, none.0___;
> 30: ADD temp[5].y, input[1]._w__, -temp[2]._w__;
> 31: CMP temp[1].y, temp[5]._y__, none._1__, none._0__;
> 32: MOV output[0].w, temp[1].___w;
> 33: KIL -temp[1].xxxx;
> 34: KIL -temp[1].yyyy;
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: RCP temp[3].w, input[0].w___;
>  1: MUL temp[3].xy, input[0].xy__, temp[3].ww__;
>  2: MAD temp[3].xy, temp[3].xy__, const[7].xy__, const[8].xy__;
>  3: MOV temp[0].x, temp[3].x___;
>  4: ADD temp[0].y, const[5]._y__, -temp[3]._y__;
>  5: TEX temp[1], input[4].xy__, 2D[1];
>  6: MUL temp[1].xyz, temp[1].xyz_, const[6].xxx_;
>  7: ADD temp[1].xyz, temp[1].xyz_, -none.111_;
>  8: MUL temp[2].xyz, input[6].xyz_, temp[1].yyy_;
>  9: MAD temp[2].xyz, input[5].xyz_, -temp[1].xxx_, temp[2].xyz_;
> 10: MAD temp[1].xyz, input[7].xyz_, temp[1].zzz_, temp[2].xyz_;
> 11: DP3 temp[2].x, temp[1].xyz_, temp[1].xyz_;
> 12: RSQ temp[2].x, |temp[2].x___|;
> 13: MUL temp[2].xyz, temp[2].xxx_, temp[1].xyz_;
> 14: TEX temp[6].xyz, temp[2].xyz_, CUBE[2];
> 15: DP3 temp[2].x, temp[2].xyz_, const[1].xyz_;
> 16: MAX temp[1].w, temp[2].___x, temp[1].___w;
> 17: MUL temp[1].xyz, temp[6].xyz_, const[2].xxx_;
> 18: MAD temp[2].xyz, temp[1].www_, const[3].xyz_, temp[1].xyz_;
> 19: TEX temp[1], input[3].xy__, 2D[1];
> 20: MUL temp[1].xyz, temp[1].xyz_, temp[2].xyz_;
> 21: ADD temp[2].xyz, -temp[1].xyz_, const[4].xyz_;
> 22: MUL temp[2].xyz, input[2].xxx_, temp[2].xyz_;
> 23: ADD output[0].xyz, temp[1].xyz_, temp[2].xyz_;
> 24: MAD temp[1].xy, temp[0].xy__, const[6].ww__, const[6].ww__;
> 25: TEX temp[7].w, temp[1].xy__, 2D[0];
> 26: ADD temp[4].x, temp[1].w___, -none.H___;
> 27: CMP temp[1].x, temp[4].x___, none.1___, none.0___;
> 28: ADD temp[5].y, input[1]._w__, -temp[7]._w__;
> 29: CMP temp[1].y, temp[5]._y__, none._1__, none._0__;
> 30: MOV output[0].w, temp[1].___w;
> 31: KIL -temp[1].xxxx;
> 32: KIL -temp[1].yyyy;
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: RCP temp[3].w, input[0].w___;
>  1: MUL temp[3].xy, input[0].xy__, temp[3].ww__;
>  2: MAD temp[3].xy, temp[3].xy__, const[7].xy__, const[8].xy__;
>  3: MOV temp[0].x, temp[3].x___;
>  4: ADD temp[0].y, const[5]._y__, -temp[3]._y__;
>  5: TEX temp[1], input[4].xy__, 2D[1];
>  6: MUL temp[1].xyz, temp[1].xyz_, const[6].xxx_;
>  7: ADD temp[1].xyz, temp[1].xyz_, -none.111_;
>  8: MUL temp[2].xyz, input[6].xyz_, temp[1].yyy_;
>  9: MAD temp[2].xyz, input[5].xyz_, -temp[1].xxx_, temp[2].xyz_;
> 10: MAD temp[1].xyz, input[7].xyz_, temp[1].zzz_, temp[2].xyz_;
> 11: DP3 temp[2].x, temp[1].xyz_, temp[1].xyz_;
> 12: RSQ temp[2].x, |temp[2].x___|;
> 13: MUL temp[2].xyz, temp[2].xxx_, temp[1].xyz_;
> 14: TEX temp[6].xyz, temp[2].xyz_, CUBE[2];
> 15: DP3 temp[2].x, temp[2].xyz_, const[1].xyz_;
> 16: MAX temp[1].w, temp[2].___x, temp[1].___w;
> 17: MUL temp[1].xyz, temp[6].xyz_, const[2].xxx_;
> 18: MAD temp[2].xyz, temp[1].www_, const[3].xyz_, temp[1].xyz_;
> 19: TEX temp[1], input[3].xy__, 2D[1];
> 20: MUL temp[1].xyz, temp[1].xyz_, temp[2].xyz_;
> 21: ADD temp[2].xyz, -temp[1].xyz_, const[4].xyz_;
> 22: MUL temp[2].xyz, input[2].xxx_, temp[2].xyz_;
> 23: ADD output[0].xyz, temp[1].xyz_, temp[2].xyz_;
> 24: MAD temp[1].xy, temp[0].xy__, const[6].ww__, const[6].ww__;
> 25: TEX temp[7].w, temp[1].xy__, 2D[0];
> 26: ADD temp[4].x, temp[1].w___, -none.H___;
> 27: CMP temp[1].x, temp[4].x___, none.1___, none.0___;
> 28: ADD temp[5].y, input[1]._w__, -temp[7]._w__;
> 29: CMP temp[1].y, temp[5]._y__, none._1__, none._0__;
> 30: MOV output[0].w, temp[1].___w;
> 31: MOV temp[8], -temp[1].xxxx;
> 32: KIL temp[8];
> 33: MOV temp[9], -temp[1].yyyy;
> 34: KIL temp[9];
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: src0.w = input[0]
>     RCP temp[3].w, src0.w
>  1: src0.xyz = input[0], src0.w = temp[3]
>     MAD temp[3].xy, src0.xy_, src0.ww_, src0.000
>  2: src0.xyz = temp[3], src1.xyz = const[7], src2.xyz = const[8]
>     MAD temp[3].xy, src0.xy_, src1.xy_, src2.xy_
>  3: src0.xyz = temp[3]
>     MAD temp[0].x, src0.x__, src0.111, src0.000
>  4: src0.xyz = const[5], src1.xyz = temp[3]
>     MAD temp[0].y, src0._y_, src0.111, -src1._y_
>  5: TEX temp[1], input[4].xy__, 2D[1];
>  6: src0.xyz = temp[1], src1.xyz = const[6]
>     MAD temp[1].xyz, src0.xyz, src1.xxx, src0.000
>  7: src0.xyz = temp[1]
>     MAD temp[1].xyz, src0.xyz, src0.111, -src0.111
>  8: src0.xyz = input[6], src1.xyz = temp[1]
>     MAD temp[2].xyz, src0.xyz, src1.yyy, src0.000
>  9: src0.xyz = input[5], src1.xyz = temp[1], src2.xyz = temp[2]
>     MAD temp[2].xyz, src0.xyz, -src1.xxx, src2.xyz
> 10: src0.xyz = input[7], src1.xyz = temp[1], src2.xyz = temp[2]
>     MAD temp[1].xyz, src0.xyz, src1.zzz, src2.xyz
> 11: src0.xyz = temp[1]
>     DP3 temp[2].x, src0.xyz, src0.xyz
> 12: src0.xyz = temp[2]
>     REPL_ALPHA temp[2].x
>     RSQ, |src0.x|
> 13: src0.xyz = temp[2], src1.xyz = temp[1]
>     MAD temp[2].xyz, src0.xxx, src1.xyz, src0.000
> 14: TEX temp[6].xyz, temp[2].xyz_, CUBE[2];
> 15: src0.xyz = temp[2], src1.xyz = const[1]
>     DP3 temp[2].x, src0.xyz, src1.xyz
> 16: src0.xyz = temp[2], src0.w = temp[1]
>     MAX temp[1].w, src0.x, src0.w
> 17: src0.xyz = temp[6], src1.xyz = const[2]
>     MAD temp[1].xyz, src0.xyz, src1.xxx, src0.000
> 18: src0.xyz = const[3], src0.w = temp[1], src1.xyz = temp[1]
>     MAD temp[2].xyz, src0.www, src0.xyz, src1.xyz
> 19: TEX temp[1], input[3].xy__, 2D[1];
> 20: src0.xyz = temp[1], src1.xyz = temp[2]
>     MAD temp[1].xyz, src0.xyz, src1.xyz, src0.000
> 21: src0.xyz = temp[1], src1.xyz = const[4]
>     MAD temp[2].xyz, -src0.xyz, src0.111, src1.xyz
> 22: src0.xyz = input[2], src1.xyz = temp[2]
>     MAD temp[2].xyz, src0.xxx, src1.xyz, src0.000
> 23: src0.xyz = temp[1], src1.xyz = temp[2]
>     MAD color[0].xyz, src0.xyz, src0.111, src1.xyz
> 24: src0.xyz = temp[0], src0.w = const[6]
>     MAD temp[1].xy, src0.xy_, src0.ww_, src0.ww_
> 25: TEX temp[7].w, temp[1].xy__, 2D[0];
> 26: src0.w = temp[1]
>     MAD temp[4].x, src0.w__, src0.111, -src0.H__
> 27: src0.xyz = temp[4]
>     CMP temp[1].x, src0.0__, src0.1__, src0.x__
> 28: src0.w = input[1], src1.w = temp[7]
>     MAD temp[5].y, src0._w_, src0.111, -src1._w_
> 29: src0.xyz = temp[5]
>     CMP temp[1].y, src0._0_, src0._1_, src0._y_
> 30: src0.w = temp[1]
>     MAD color[0].w, src0.w, src0.1, src0.0
> 31: src0.xyz = temp[1]
>     MAD temp[8].xyz, -src0.xxx, src0.111, src0.000
>     MAD temp[8].w, -src0.x, src0.1, src0.0
> 32: KIL temp[8];
> 33: src0.xyz = temp[1]
>     MAD temp[9].xyz, -src0.yyy, src0.111, src0.000
>     MAD temp[9].w, -src0.y, src0.1, src0.0
> 34: KIL temp[9];
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[1], input[4].xy__, 2D[1];
>  2: src0.xyz = temp[1], src0.w = input[0], src1.xyz = const[6]
>     MAD temp[1].xyz, src0.xyz, src1.xxx, src0.000
>     RCP temp[3].w, src0.w
>  3: src0.xyz = input[0], src0.w = temp[3]
>     MAD temp[3].xy, src0.xy_, src0.ww_, src0.000
>  4: src0.xyz = temp[3], src1.xyz = const[7], src2.xyz = const[8]
>     MAD temp[3].xy, src0.xy_, src1.xy_, src2.xy_
>  5: src0.xyz = const[5], src1.xyz = temp[3]
>     MAD temp[0].y, src0._y_, src0.111, -src1._y_
>  6: src0.xyz = temp[3]
>     MAD temp[0].x, src0.x__, src0.111, src0.000
>  7: src0.xyz = temp[1]
>     MAD temp[1].xyz, src0.xyz, src0.111, -src0.111
>  8: src0.xyz = input[6], src1.xyz = temp[1]
>     MAD temp[2].xyz, src0.xyz, src1.yyy, src0.000
>  9: src0.xyz = input[5], src1.xyz = temp[1], src2.xyz = temp[2]
>     MAD temp[2].xyz, src0.xyz, -src1.xxx, src2.xyz
> 10: src0.xyz = input[7], src1.xyz = temp[1], src2.xyz = temp[2]
>     MAD temp[1].xyz, src0.xyz, src1.zzz, src2.xyz
> 11: src0.xyz = temp[1]
>     DP3 temp[2].x, src0.xyz, src0.xyz
> 12: src0.xyz = temp[2]
>     REPL_ALPHA temp[2].x
>     RSQ, |src0.x|
> 13: src0.xyz = temp[2], src1.xyz = temp[1]
>     MAD temp[2].xyz, src0.xxx, src1.xyz, src0.000
> 14: BEGIN_TEX;
> 15: TEX temp[6].xyz, temp[2].xyz_, CUBE[2];
> 16: src0.xyz = temp[6], src1.xyz = const[2]
>     MAD temp[1].xyz, src0.xyz, src1.xxx, src0.000
> 17: src0.xyz = temp[2], src1.xyz = const[1]
>     DP3 temp[2].x, src0.xyz, src1.xyz
> 18: src0.xyz = temp[2], src0.w = temp[1]
>     MAX temp[1].w, src0.x, src0.w
> 19: src0.xyz = const[3], src0.w = temp[1], src1.xyz = temp[1]
>     MAD temp[2].xyz, src0.www, src0.xyz, src1.xyz
> 20: BEGIN_TEX;
> 21: TEX temp[1], input[3].xy__, 2D[1];
> 22: src0.w = temp[1]
>     MAD temp[4].x, src0.w__, src0.111, -src0.H__
>     MAD color[0].w, src0.w, src0.1, src0.0
> 23: src0.xyz = temp[1], src1.xyz = temp[2]
>     MAD temp[1].xyz, src0.xyz, src1.xyz, src0.000
> 24: src0.xyz = temp[1], src1.xyz = const[4]
>     MAD temp[2].xyz, -src0.xyz, src0.111, src1.xyz
> 25: src0.xyz = input[2], src1.xyz = temp[2]
>     MAD temp[2].xyz, src0.xxx, src1.xyz, src0.000
> 26: src0.xyz = temp[1], src1.xyz = temp[2]
>     MAD color[0].xyz, src0.xyz, src0.111, src1.xyz
> 27: src0.xyz = temp[0], src0.w = const[6]
>     MAD temp[1].xy, src0.xy_, src0.ww_, src0.ww_
> 28: BEGIN_TEX;
> 29: TEX temp[7].w, temp[1].xy__, 2D[0];
> 30: src0.w = input[1], src1.w = temp[7]
>     MAD temp[5].y, src0._w_, src0.111, -src1._w_
> 31: src0.xyz = temp[5]
>     CMP temp[1].y, src0._0_, src0._1_, src0._y_
> 32: src0.xyz = temp[1]
>     MAD temp[9].xyz, -src0.yyy, src0.111, src0.000
>     MAD temp[9].w, -src0.y, src0.1, src0.0
> 33: src0.xyz = temp[4]
>     CMP temp[1].x, src0.0__, src0.1__, src0.x__
> 34: src0.xyz = temp[1]
>     MAD temp[8].xyz, -src0.xxx, src0.111, src0.000
>     MAD temp[8].w, -src0.x, src0.1, src0.0
> 35: BEGIN_TEX;
> 36: KIL temp[8];
> 37: KIL temp[9];
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[8], temp[3].xy__, 2D[1];
>  2: src0.xyz = temp[8], src0.w = temp[7], src1.xyz = const[6]
>     MAD temp[8].xyz, src0.xyz, src1.xxx, src0.000
>     RCP temp[9].w, src0.w
>  3: src0.xyz = temp[7], src0.w = temp[9]
>     MAD temp[9].xy, src0.xy_, src0.ww_, src0.000
>  4: src0.xyz = temp[9], src1.xyz = const[7], src2.xyz = const[8]
>     MAD temp[9].xy, src0.xy_, src1.xy_, src2.xy_
>  5: src0.xyz = const[5], src1.xyz = temp[9]
>     MAD temp[3].y, src0._y_, src0.111, -src1._y_
>  6: src0.xyz = temp[9]
>     MAD temp[3].x, src0.x__, src0.111, src0.000
>  7: src0.xyz = temp[8]
>     MAD temp[8].xyz, src0.xyz, src0.111, -src0.111
>  8: src0.xyz = temp[5], src1.xyz = temp[8]
>     MAD temp[5].xyz, src0.xyz, src1.yyy, src0.000
>  9: src0.xyz = temp[4], src1.xyz = temp[8], src2.xyz = temp[5]
>     MAD temp[5].xyz, src0.xyz, -src1.xxx, src2.xyz
> 10: src0.xyz = temp[6], src1.xyz = temp[8], src2.xyz = temp[5]
>     MAD temp[8].xyz, src0.xyz, src1.zzz, src2.xyz
> 11: src0.xyz = temp[8]
>     DP3 temp[5].x, src0.xyz, src0.xyz
> 12: src0.xyz = temp[5]
>     REPL_ALPHA temp[5].x
>     RSQ, |src0.x|
> 13: src0.xyz = temp[5], src1.xyz = temp[8]
>     MAD temp[5].xyz, src0.xxx, src1.xyz, src0.000
> 14: BEGIN_TEX;
> 15: TEX temp[4].xyz, temp[5].xyz_, CUBE[2];
> 16: src0.xyz = temp[4], src1.xyz = const[2]
>     MAD temp[8].xyz, src0.xyz, src1.xxx, src0.000
> 17: src0.xyz = temp[5], src1.xyz = const[1]
>     DP3 temp[5].x, src0.xyz, src1.xyz
> 18: src0.xyz = temp[5], src0.w = temp[8]
>     MAX temp[8].w, src0.x, src0.w
> 19: src0.xyz = const[3], src0.w = temp[8], src1.xyz = temp[8]
>     MAD temp[5].xyz, src0.www, src0.xyz, src1.xyz
> 20: BEGIN_TEX;
> 21: TEX temp[8], temp[2].xy__, 2D[1];
> 22: src0.w = temp[8]
>     MAD temp[2].x, src0.w__, src0.111, -src0.H__
>     MAD color[0].w, src0.w, src0.1, src0.0
> 23: src0.xyz = temp[8], src1.xyz = temp[5]
>     MAD temp[8].xyz, src0.xyz, src1.xyz, src0.000
> 24: src0.xyz = temp[8], src1.xyz = const[4]
>     MAD temp[5].xyz, -src0.xyz, src0.111, src1.xyz
> 25: src0.xyz = temp[1], src1.xyz = temp[5]
>     MAD temp[5].xyz, src0.xxx, src1.xyz, src0.000
> 26: src0.xyz = temp[8], src1.xyz = temp[5]
>     MAD color[0].xyz, src0.xyz, src0.111, src1.xyz
> 27: src0.xyz = temp[3], src0.w = const[6]
>     MAD temp[8].xy, src0.xy_, src0.ww_, src0.ww_
> 28: BEGIN_TEX;
> 29: TEX temp[1].w, temp[8].xy__, 2D[0];
> 30: src0.w = temp[0], src1.w = temp[1]
>     MAD temp[0].y, src0._w_, src0.111, -src1._w_
> 31: src0.xyz = temp[0]
>     CMP temp[8].y, src0._0_, src0._1_, src0._y_
> 32: src0.xyz = temp[8]
>     MAD temp[1].xyz, -src0.yyy, src0.111, src0.000
>     MAD temp[1].w, -src0.y, src0.1, src0.0
> 33: src0.xyz = temp[2]
>     CMP temp[8].x, src0.0__, src0.1__, src0.x__
> 34: src0.xyz = temp[8]
>     MAD temp[0].xyz, -src0.xxx, src0.111, src0.000
>     MAD temp[0].w, -src0.x, src0.1, src0.0
> 35: BEGIN_TEX;
> 36: KIL temp[0];
> 37: KIL temp[1];
>r300compiler error: r300_fragprog_emit.c::begin_tex(): Too many texture indirections
>pc=38*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 0, tex_end: 0  (code_addr: 00000000)
>  TEX:
>    TEX t8, t3, texture[1]   (00008a03)
>  0: xyz:  t8  c6  t0 -> t8.xyz               (03a00988)
>       w:  t7  t0  t0 -> t9.w                 (00a40007)
>     xyz:   t8.xyz   c6.xxx      0.0    op: 00050280
>       w:     t7.w     t8.x     t8.x    op: 05000009
>r300 FP: Compiler Error:
>r300_fragprog_emit.c::begin_tex(): Too many texture indirections
>Using a dummy shader instead.
>If there's an 'unknown opcode' message, please file a bug report and attach this log.
>r300: Initial fragment program
>FRAG
>DCL OUT[0], COLOR
>IMM FLT32 {    0.0000,     1.0000,     0.0000,     0.0000}
>  0: MOV OUT[0], IMM[0].xxxy
>  1: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0001;
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0001;
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0001;
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0001;
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0001;
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0001;
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0001;
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0001;
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: 
>     MAD color[0].xyz, src0.000, src0.111, src0.000
>     MAD color[0].w, src0.1, src0.1, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: 
>     MAD color[0].xyz, src0.000, src0.111, src0.000
>     MAD color[0].w, src0.1, src0.1, src0.0
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: 
>     MAD color[0].xyz, src0.000, src0.111, src0.000
>     MAD color[0].w, src0.1, src0.1, src0.0
>pc=39*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 0, tex_end: 0  (code_addr: 00400000)
>  0: xyz:  t0  t0  t0 -> o0.xyz               (1c000000)
>       w:  t0  t0  t0 -> o0.w                 (01000000)
>     xyz:      0.0      1.0      0.0    op: 00050a94
>       w:      1.0      1.0      0.0    op: 00040891
>r300: Initial fragment program
>FRAG
>DCL IN[0], POSITION, LINEAR
>DCL IN[1], COLOR[1], LINEAR
>DCL IN[2], GENERIC[0], PERSPECTIVE
>DCL IN[3], GENERIC[2], PERSPECTIVE
>DCL IN[4], GENERIC[3], PERSPECTIVE
>DCL IN[5], GENERIC[4], PERSPECTIVE
>DCL IN[6], GENERIC[5], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL SAMP[1]
>DCL SAMP[2]
>DCL SAMP[3]
>DCL CONST[5..6]
>DCL CONST[0]
>DCL CONST[2]
>DCL CONST[4]
>DCL TEMP[0..10]
>IMM FLT32 {    2.0000,     1.0000,     0.0001,     0.0000}
>IMM FLT32 {    0.1250,     0.0000,     0.0000,     0.0000}
>  0: MOV TEMP[0], IN[0]
>  1: SUB TEMP[0].y, CONST[6], IN[0]
>  2: TEX TEMP[1], IN[2], SAMP[0], 2D
>  3: TEX TEMP[2], IN[2], SAMP[1], 2D
>  4: LRP TEMP[3], CONST[0], TEMP[2], TEMP[1]
>  5: MOV TEMP[3].x, TEMP[3].wwww
>  6: MAD TEMP[3], TEMP[3], IMM[0].xxxx, -IMM[0].yyyy
>  7: MUL TEMP[3].xy, TEMP[3], CONST[2].zzzz
>  8: MUL_SAT TEMP[1].w, IN[3].wwww, IMM[0].zzzz
>  9: SUB TEMP[2], IMM[0].wwww, TEMP[3]
> 10: MAD TEMP[3], TEMP[3], TEMP[1].wwww, TEMP[2]
> 11: MOV TEMP[3].z, IMM[0].wwww
> 12: DP3 TEMP[4].x, TEMP[3], TEMP[3]
> 13: ADD TEMP[4].x, IMM[0].yyyy, -TEMP[4].xxxx
> 14: RSQ TEMP[4].x, TEMP[4].xxxx
> 15: RCP TEMP[3].z, TEMP[4].xxxx
> 16: DP3 TEMP[2].x, TEMP[3], IN[4]
> 17: DP3 TEMP[2].y, TEMP[3], IN[5]
> 18: DP3 TEMP[2].z, TEMP[3], IN[6]
> 19: MOV TEMP[5], TEMP[0]
> 20: MOV TEMP[6], IN[3]
> 21: DP3 TEMP[3], TEMP[6], TEMP[2]
> 22: MUL TEMP[3], TEMP[3], IMM[0].xxxx
> 23: MAD TEMP[7], -TEMP[2], TEMP[3].xxxx, TEMP[6]
> 24: MOV TEMP[8], IMM[1].xxxx
> 25: TEX TEMP[2], TEMP[5], SAMP[2], RECT
> 26: MUL TEMP[2], TEMP[2], CONST[4]
> 27: TEX TEMP[3], TEMP[7], SAMP[3], CUBE
> 28: SUB TEMP[5], TEMP[3], TEMP[2]
> 29: MAD TEMP[5], TEMP[5], TEMP[8], TEMP[2]
> 30: MOV TEMP[9], TEMP[5]
> 31: ADD OUT[0].w, TEMP[10].xxxx, TEMP[10].yyyy
> 32: LRP OUT[0].xyz, IN[1].xxxx, CONST[5], TEMP[9]
> 33: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: RCP temp[11].w, input[0].wwww;
>  1: MUL temp[11].xyz, input[0], temp[11].wwww;
>  2: MAD temp[11].xyz, temp[11].xyz0, const[9].xyz0, const[10].xyz0;
>  3: MOV temp[0], temp[11];
>  4: SUB temp[0].y, const[6], temp[11];
>  5: TEX temp[1], input[2], 2D[0];
>  6: TEX temp[2], input[2], 2D[1];
>  7: LRP temp[3], const[0], temp[2], temp[1];
>  8: MOV temp[3].x, temp[3].wwww;
>  9: MAD temp[3], temp[3], const[7].xxxx, -const[7].yyyy;
> 10: MUL temp[3].xy, temp[3], const[2].zzzz;
> 11: MUL_SAT temp[1].w, input[3].wwww, const[7].zzzz;
> 12: SUB temp[2], const[7].wwww, temp[3];
> 13: MAD temp[3], temp[3], temp[1].wwww, temp[2];
> 14: MOV temp[3].z, const[7].wwww;
> 15: DP3 temp[4].x, temp[3], temp[3];
> 16: ADD temp[4].x, const[7].yyyy, -temp[4].xxxx;
> 17: RSQ temp[4].x, temp[4].xxxx;
> 18: RCP temp[3].z, temp[4].xxxx;
> 19: DP3 temp[2].x, temp[3], input[4];
> 20: DP3 temp[2].y, temp[3], input[5];
> 21: DP3 temp[2].z, temp[3], input[6];
> 22: MOV temp[5], temp[0];
> 23: MOV temp[6], input[3];
> 24: DP3 temp[3], temp[6], temp[2];
> 25: MUL temp[3], temp[3], const[7].xxxx;
> 26: MAD temp[7], -temp[2], temp[3].xxxx, temp[6];
> 27: MOV temp[8], const[8].xxxx;
> 28: TEX temp[2], temp[5], RECT[2];
> 29: MUL temp[2], temp[2], const[4];
> 30: TEX temp[3], temp[7], CUBE[3];
> 31: SUB temp[5], temp[3], temp[2];
> 32: MAD temp[5], temp[5], temp[8], temp[2];
> 33: MOV temp[9], temp[5];
> 34: ADD output[0].w, temp[10].xxxx, temp[10].yyyy;
> 35: LRP output[0].xyz, input[1].xxxx, const[5], temp[9];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: RCP temp[11].w, input[0].wwww;
>  1: MUL temp[11].xyz, input[0], temp[11].wwww;
>  2: MAD temp[11].xyz, temp[11].xyz0, const[9].xyz0, const[10].xyz0;
>  3: MOV temp[0], temp[11];
>  4: SUB temp[0].y, const[6], temp[11];
>  5: TEX temp[1], input[2], 2D[0];
>  6: TEX temp[2], input[2], 2D[1];
>  7: LRP temp[3], const[0], temp[2], temp[1];
>  8: MOV temp[3].x, temp[3].wwww;
>  9: MAD temp[3], temp[3], const[7].xxxx, -const[7].yyyy;
> 10: MUL temp[3].xy, temp[3], const[2].zzzz;
> 11: MUL_SAT temp[1].w, input[3].wwww, const[7].zzzz;
> 12: SUB temp[2], const[7].wwww, temp[3];
> 13: MAD temp[3], temp[3], temp[1].wwww, temp[2];
> 14: MOV temp[3].z, const[7].wwww;
> 15: DP3 temp[4].x, temp[3], temp[3];
> 16: ADD temp[4].x, const[7].yyyy, -temp[4].xxxx;
> 17: RSQ temp[4].x, temp[4].xxxx;
> 18: RCP temp[3].z, temp[4].xxxx;
> 19: DP3 temp[2].x, temp[3], input[4];
> 20: DP3 temp[2].y, temp[3], input[5];
> 21: DP3 temp[2].z, temp[3], input[6];
> 22: MOV temp[5], temp[0];
> 23: MOV temp[6], input[3];
> 24: DP3 temp[3], temp[6], temp[2];
> 25: MUL temp[3], temp[3], const[7].xxxx;
> 26: MAD temp[7], -temp[2], temp[3].xxxx, temp[6];
> 27: MOV temp[8], const[8].xxxx;
> 28: TEX temp[2], temp[5], RECT[2];
> 29: MUL temp[2], temp[2], const[4];
> 30: TEX temp[3], temp[7], CUBE[3];
> 31: SUB temp[5], temp[3], temp[2];
> 32: MAD temp[5], temp[5], temp[8], temp[2];
> 33: MOV temp[9], temp[5];
> 34: ADD output[0].w, temp[10].xxxx, temp[10].yyyy;
> 35: LRP output[0].xyz, input[1].xxxx, const[5], temp[9];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: RCP temp[11].w, input[0].wwww;
>  1: MUL temp[11].xyz, input[0], temp[11].wwww;
>  2: MAD temp[11].xyz, temp[11].xyz0, const[9].xyz0, const[10].xyz0;
>  3: MOV temp[0], temp[11];
>  4: SUB temp[0].y, const[6], temp[11];
>  5: TEX temp[1], input[2], 2D[0];
>  6: TEX temp[2], input[2], 2D[1];
>  7: LRP temp[3], const[0], temp[2], temp[1];
>  8: MOV temp[3].x, temp[3].wwww;
>  9: MAD temp[3], temp[3], const[7].xxxx, -const[7].yyyy;
> 10: MUL temp[3].xy, temp[3], const[2].zzzz;
> 11: MUL_SAT temp[1].w, input[3].wwww, const[7].zzzz;
> 12: SUB temp[2], const[7].wwww, temp[3];
> 13: MAD temp[3], temp[3], temp[1].wwww, temp[2];
> 14: MOV temp[3].z, const[7].wwww;
> 15: DP3 temp[4].x, temp[3], temp[3];
> 16: ADD temp[4].x, const[7].yyyy, -temp[4].xxxx;
> 17: RSQ temp[4].x, temp[4].xxxx;
> 18: RCP temp[3].z, temp[4].xxxx;
> 19: DP3 temp[2].x, temp[3], input[4];
> 20: DP3 temp[2].y, temp[3], input[5];
> 21: DP3 temp[2].z, temp[3], input[6];
> 22: MOV temp[5], temp[0];
> 23: MOV temp[6], input[3];
> 24: DP3 temp[3], temp[6], temp[2];
> 25: MUL temp[3], temp[3], const[7].xxxx;
> 26: MAD temp[7], -temp[2], temp[3].xxxx, temp[6];
> 27: MOV temp[8], const[8].xxxx;
> 28: TEX temp[2], temp[5], RECT[2];
> 29: MUL temp[2], temp[2], const[4];
> 30: TEX temp[3], temp[7], CUBE[3];
> 31: SUB temp[5], temp[3], temp[2];
> 32: MAD temp[5], temp[5], temp[8], temp[2];
> 33: MOV temp[9], temp[5];
> 34: ADD output[0].w, temp[10].xxxx, temp[10].yyyy;
> 35: LRP output[0].xyz, input[1].xxxx, const[5], temp[9];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: RCP temp[11].w, input[0].wwww;
>  1: MUL temp[11].xyz, input[0], temp[11].wwww;
>  2: MAD temp[11].xyz, temp[11].xyz0, const[9].xyz0, const[10].xyz0;
>  3: MOV temp[0], temp[11];
>  4: ADD temp[0].y, const[6], -temp[11];
>  5: TEX temp[1], input[2], 2D[0];
>  6: TEX temp[2], input[2], 2D[1];
>  7: ADD temp[12], temp[2], -temp[1];
>  8: MAD temp[3], const[0], temp[12], temp[1];
>  9: MOV temp[3].x, temp[3].wwww;
> 10: MAD temp[3], temp[3], const[7].xxxx, -const[7].yyyy;
> 11: MUL temp[3].xy, temp[3], const[2].zzzz;
> 12: MUL_SAT temp[1].w, input[3].wwww, const[7].zzzz;
> 13: ADD temp[2], const[7].wwww, -temp[3];
> 14: MAD temp[3], temp[3], temp[1].wwww, temp[2];
> 15: MOV temp[3].z, const[7].wwww;
> 16: DP3 temp[4].x, temp[3], temp[3];
> 17: ADD temp[4].x, const[7].yyyy, -temp[4].xxxx;
> 18: RSQ temp[4].x, |temp[4].xxxx|;
> 19: RCP temp[3].z, temp[4].xxxx;
> 20: DP3 temp[2].x, temp[3], input[4];
> 21: DP3 temp[2].y, temp[3], input[5];
> 22: DP3 temp[2].z, temp[3], input[6];
> 23: MOV temp[5], temp[0];
> 24: MOV temp[6], input[3];
> 25: DP3 temp[3], temp[6], temp[2];
> 26: MUL temp[3], temp[3], const[7].xxxx;
> 27: MAD temp[7], -temp[2], temp[3].xxxx, temp[6];
> 28: MOV temp[8], const[8].xxxx;
> 29: TEX temp[2], temp[5], RECT[2];
> 30: MUL temp[2], temp[2], const[4];
> 31: TEX temp[3], temp[7], CUBE[3];
> 32: ADD temp[5], temp[3], -temp[2];
> 33: MAD temp[5], temp[5], temp[8], temp[2];
> 34: MOV temp[9], temp[5];
> 35: ADD output[0].w, temp[10].xxxx, temp[10].yyyy;
> 36: ADD temp[13], const[5], -temp[9];
> 37: MAD output[0].xyz, input[1].xxxx, temp[13], temp[9];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: RCP temp[11].w, input[0].wwww;
>  1: MUL temp[11].xyz, input[0], temp[11].wwww;
>  2: MAD temp[11].xyz, temp[11].xyz0, const[9].xyz0, const[10].xyz0;
>  3: MOV temp[0], temp[11];
>  4: ADD temp[0].y, const[6], -temp[11];
>  5: TEX temp[1], input[2], 2D[0];
>  6: TEX temp[2], input[2], 2D[1];
>  7: ADD temp[12], temp[2], -temp[1];
>  8: MAD temp[3], const[0], temp[12], temp[1];
>  9: MOV temp[3].x, temp[3].wwww;
> 10: MAD temp[3], temp[3], const[7].xxxx, -const[7].yyyy;
> 11: MUL temp[3].xy, temp[3], const[2].zzzz;
> 12: MUL_SAT temp[1].w, input[3].wwww, const[7].zzzz;
> 13: ADD temp[2], const[7].wwww, -temp[3];
> 14: MAD temp[3], temp[3], temp[1].wwww, temp[2];
> 15: MOV temp[3].z, const[7].wwww;
> 16: DP3 temp[4].x, temp[3], temp[3];
> 17: ADD temp[4].x, const[7].yyyy, -temp[4].xxxx;
> 18: RSQ temp[4].x, |temp[4].xxxx|;
> 19: RCP temp[3].z, temp[4].xxxx;
> 20: DP3 temp[2].x, temp[3], input[4];
> 21: DP3 temp[2].y, temp[3], input[5];
> 22: DP3 temp[2].z, temp[3], input[6];
> 23: MOV temp[5], temp[0];
> 24: MOV temp[6], input[3];
> 25: DP3 temp[3], temp[6], temp[2];
> 26: MUL temp[3], temp[3], const[7].xxxx;
> 27: MAD temp[7], -temp[2], temp[3].xxxx, temp[6];
> 28: MOV temp[8], const[8].xxxx;
> 29: MUL temp[14], temp[5], const[11];
> 30: TEX temp[2], temp[14], 2D[2];
> 31: MUL temp[2], temp[2], const[4];
> 32: TEX temp[3], temp[7], CUBE[3];
> 33: ADD temp[5], temp[3], -temp[2];
> 34: MAD temp[5], temp[5], temp[8], temp[2];
> 35: MOV temp[9], temp[5];
> 36: ADD output[0].w, temp[10].xxxx, temp[10].yyyy;
> 37: ADD temp[13], const[5], -temp[9];
> 38: MAD output[0].xyz, input[1].xxxx, temp[13], temp[9];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: RCP temp[11].w, input[0].w___;
>  1: MUL temp[11].xy, input[0].xy__, temp[11].ww__;
>  2: MAD temp[11].xy, temp[11].xy__, const[9].xy__, const[10].xy__;
>  3: MOV temp[0].x, temp[11].x___;
>  4: ADD temp[0].y, const[6]._y__, -temp[11]._y__;
>  5: TEX temp[1].yw, input[2].xy__, 2D[0];
>  6: TEX temp[2].yw, input[2].xy__, 2D[1];
>  7: ADD temp[12].yw, temp[2]._y_w, -temp[1]._y_w;
>  8: MAD temp[3].yw, const[0]._y_w, temp[12]._y_w, temp[1]._y_w;
>  9: MOV temp[3].x, temp[3].w___;
> 10: MAD temp[3].xy, temp[3].xy__, const[7].xx__, -const[7].yy__;
> 11: MUL temp[3].xy, temp[3].xy__, const[2].zz__;
> 12: MUL_SAT temp[1].w, input[3].___w, const[7].___z;
> 13: ADD temp[2].xy, const[7].ww__, -temp[3].xy__;
> 14: MAD temp[3].xy, temp[3].xy__, temp[1].ww__, temp[2].xy__;
> 15: MOV temp[3].z, const[7].__w_;
> 16: DP3 temp[4].x, temp[3].xyz_, temp[3].xyz_;
> 17: ADD temp[4].x, const[7].y___, -temp[4].x___;
> 18: RSQ temp[4].x, |temp[4].x___|;
> 19: RCP temp[3].z, temp[4].x___;
> 20: DP3 temp[2].x, temp[3].xyz_, input[4].xyz_;
> 21: DP3 temp[2].y, temp[3].xyz_, input[5].xyz_;
> 22: DP3 temp[2].z, temp[3].xyz_, input[6].xyz_;
> 23: MOV temp[5].xy, temp[0].xy__;
> 24: MOV temp[6].xyz, input[3].xyz_;
> 25: DP3 temp[3].x, temp[6].xyz_, temp[2].xyz_;
> 26: MUL temp[3].x, temp[3].x___, const[7].x___;
> 27: MAD temp[7].xyz, -temp[2].xyz_, temp[3].xxx_, temp[6].xyz_;
> 28: MOV temp[8].xyz, const[8].xxx_;
> 29: MUL temp[14].xy, temp[5].xy__, const[11].xy__;
> 30: TEX temp[2].xyz, temp[14].xy__, 2D[2];
> 31: MUL temp[2].xyz, temp[2].xyz_, const[4].xyz_;
> 32: TEX temp[3].xyz, temp[7].xyz_, CUBE[3];
> 33: ADD temp[5].xyz, temp[3].xyz_, -temp[2].xyz_;
> 34: MAD temp[5].xyz, temp[5].xyz_, temp[8].xyz_, temp[2].xyz_;
> 35: MOV temp[9].xyz, temp[5].xyz_;
> 36: ADD output[0].w, temp[10].___x, temp[10].___y;
> 37: ADD temp[13].xyz, const[5].xyz_, -temp[9].xyz_;
> 38: MAD output[0].xyz, input[1].xxx_, temp[13].xyz_, temp[9].xyz_;
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: RCP temp[11].w, input[0].w___;
>  1: MUL temp[11].xy, input[0].xy__, temp[11].ww__;
>  2: MAD temp[11].xy, temp[11].xy__, const[9].xy__, const[10].xy__;
>  3: MOV temp[0].x, temp[11].x___;
>  4: ADD temp[0].y, const[6]._y__, -temp[11]._y__;
>  5: TEX temp[1].yw, input[2].xy__, 2D[0];
>  6: TEX temp[2].yw, input[2].xy__, 2D[1];
>  7: ADD temp[12].yw, temp[2]._y_w, -temp[1]._y_w;
>  8: MAD temp[3].yw, const[0]._y_w, temp[12]._y_w, temp[1]._y_w;
>  9: MOV temp[3].x, temp[3].w___;
> 10: MAD temp[3].xy, temp[3].xy__, const[7].xx__, -none.11__;
> 11: MUL temp[3].xy, temp[3].xy__, const[2].zz__;
> 12: MUL_SAT temp[1].w, input[3].___w, const[7].___z;
> 13: MAD temp[3].xy, temp[3].xy__, temp[1].ww__, temp[3].-x-y__;
> 14: MOV temp[3].z, none.__0_;
> 15: DP3 temp[4].x, temp[3].xyz_, temp[3].xyz_;
> 16: ADD temp[4].x, none.1___, -temp[4].x___;
> 17: RSQ temp[4].x, |temp[4].x___|;
> 18: RCP temp[3].z, temp[4].x___;
> 19: DP3 temp[2].x, temp[3].xyz_, input[4].xyz_;
> 20: DP3 temp[2].y, temp[3].xyz_, input[5].xyz_;
> 21: DP3 temp[2].z, temp[3].xyz_, input[6].xyz_;
> 22: DP3 temp[3].x, input[3].xyz_, temp[2].xyz_;
> 23: MUL temp[3].x, temp[3].x___, const[7].x___;
> 24: MAD temp[7].xyz, -temp[2].xyz_, temp[3].xxx_, input[3].xyz_;
> 25: MUL temp[14].xy, temp[0].xy__, const[11].xy__;
> 26: TEX temp[2].xyz, temp[14].xy__, 2D[2];
> 27: MUL temp[2].xyz, temp[2].xyz_, const[4].xyz_;
> 28: TEX temp[3].xyz, temp[7].xyz_, CUBE[3];
> 29: ADD temp[5].xyz, temp[3].xyz_, -temp[2].xyz_;
> 30: MAD temp[5].xyz, temp[5].xyz_, const[8].xxx_, temp[2].xyz_;
> 31: ADD output[0].w, temp[10].___x, temp[10].___y;
> 32: ADD temp[13].xyz, const[5].xyz_, -temp[5].xyz_;
> 33: MAD output[0].xyz, input[1].xxx_, temp[13].xyz_, temp[5].xyz_;
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: RCP temp[11].w, input[0].w___;
>  1: MUL temp[11].xy, input[0].xy__, temp[11].ww__;
>  2: MAD temp[11].xy, temp[11].xy__, const[9].xy__, const[10].xy__;
>  3: MOV temp[0].x, temp[11].x___;
>  4: ADD temp[0].y, const[6]._y__, -temp[11]._y__;
>  5: TEX temp[1].yw, input[2].xy__, 2D[0];
>  6: TEX temp[2].yw, input[2].xy__, 2D[1];
>  7: ADD temp[12].yw, temp[2]._y_w, -temp[1]._y_w;
>  8: MAD temp[3].yw, const[0]._y_w, temp[12]._y_w, temp[1]._y_w;
>  9: MOV temp[3].x, temp[3].w___;
> 10: MAD temp[3].xy, temp[3].xy__, const[7].xx__, -none.11__;
> 11: MUL temp[3].xy, temp[3].xy__, const[2].zz__;
> 12: MUL_SAT temp[1].w, input[3].___w, const[7].___z;
> 13: MAD temp[3].xy, temp[3].xy__, temp[1].ww__, temp[3].-x-y__;
> 14: MOV temp[3].z, none.__0_;
> 15: DP3 temp[4].x, temp[3].xyz_, temp[3].xyz_;
> 16: ADD temp[4].x, none.1___, -temp[4].x___;
> 17: RSQ temp[4].x, |temp[4].x___|;
> 18: RCP temp[3].z, temp[4].x___;
> 19: DP3 temp[2].x, temp[3].xyz_, input[4].xyz_;
> 20: DP3 temp[2].y, temp[3].xyz_, input[5].xyz_;
> 21: DP3 temp[2].z, temp[3].xyz_, input[6].xyz_;
> 22: DP3 temp[3].x, input[3].xyz_, temp[2].xyz_;
> 23: MUL temp[3].x, temp[3].x___, const[7].x___;
> 24: MAD temp[7].xyz, -temp[2].xyz_, temp[3].xxx_, input[3].xyz_;
> 25: MUL temp[14].xy, temp[0].xy__, const[11].xy__;
> 26: TEX temp[2].xyz, temp[14].xy__, 2D[2];
> 27: MUL temp[2].xyz, temp[2].xyz_, const[4].xyz_;
> 28: TEX temp[3].xyz, temp[7].xyz_, CUBE[3];
> 29: ADD temp[5].xyz, temp[3].xyz_, -temp[2].xyz_;
> 30: MAD temp[5].xyz, temp[5].xyz_, const[8].xxx_, temp[2].xyz_;
> 31: ADD output[0].w, temp[10].___x, temp[10].___y;
> 32: ADD temp[13].xyz, const[5].xyz_, -temp[5].xyz_;
> 33: MAD output[0].xyz, input[1].xxx_, temp[13].xyz_, temp[5].xyz_;
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: src0.w = input[0]
>     RCP temp[11].w, src0.w
>  1: src0.xyz = input[0], src0.w = temp[11]
>     MAD temp[11].xy, src0.xy_, src0.ww_, src0.000
>  2: src0.xyz = temp[11], src1.xyz = const[9], src2.xyz = const[10]
>     MAD temp[11].xy, src0.xy_, src1.xy_, src2.xy_
>  3: src0.xyz = temp[11]
>     MAD temp[0].x, src0.x__, src0.111, src0.000
>  4: src0.xyz = const[6], src1.xyz = temp[11]
>     MAD temp[0].y, src0._y_, src0.111, -src1._y_
>  5: TEX temp[1].yw, input[2].xy__, 2D[0];
>  6: TEX temp[2].yw, input[2].xy__, 2D[1];
>  7: src0.xyz = temp[2], src0.w = temp[2], src1.xyz = temp[1], src1.w = temp[1]
>     MAD temp[12].y, src0._y_, src0.111, -src1._y_
>     MAD temp[12].w, src0.w, src0.1, -src1.w
>  8: src0.xyz = const[0], src0.w = const[0], src1.xyz = temp[12], src1.w = temp[12], src2.xyz = temp[1], src2.w = temp[1]
>     MAD temp[3].y, src0._y_, src1._y_, src2._y_
>     MAD temp[3].w, src0.w, src1.w, src2.w
>  9: src0.w = temp[3]
>     MAD temp[3].x, src0.w__, src0.111, src0.000
> 10: src0.xyz = temp[3], src1.xyz = const[7]
>     MAD temp[3].xy, src0.xy_, src1.xx_, -src0.11_
> 11: src0.xyz = temp[3], src1.xyz = const[2]
>     MAD temp[3].xy, src0.xy_, src1.zz_, src0.000
> 12: src0.xyz = const[7], src0.w = input[3]
>     MAD_SAT temp[1].w, src0.w, src0.z, src0.0
> 13: src0.xyz = temp[3], src0.w = temp[1]
>     MAD temp[3].xy, src0.xy_, src0.ww_, -src0.xy_
> 14: 
>     MAD temp[3].z, src0.__0, src0.111, src0.000
> 15: src0.xyz = temp[3]
>     DP3 temp[4].x, src0.xyz, src0.xyz
> 16: src0.xyz = temp[4]
>     MAD temp[4].x, src0.1__, src0.111, -src0.x__
> 17: src0.xyz = temp[4]
>     REPL_ALPHA temp[4].x
>     RSQ, |src0.x|
> 18: src0.xyz = temp[4]
>     REPL_ALPHA temp[3].z
>     RCP, src0.x
> 19: src0.xyz = temp[3], src1.xyz = input[4]
>     DP3 temp[2].x, src0.xyz, src1.xyz
> 20: src0.xyz = temp[3], src1.xyz = input[5]
>     DP3 temp[2].y, src0.xyz, src1.xyz
> 21: src0.xyz = temp[3], src1.xyz = input[6]
>     DP3 temp[2].z, src0.xyz, src1.xyz
> 22: src0.xyz = input[3], src1.xyz = temp[2]
>     DP3 temp[3].x, src0.xyz, src1.xyz
> 23: src0.xyz = temp[3], src1.xyz = const[7]
>     MAD temp[3].x, src0.x__, src1.x__, src0.000
> 24: src0.xyz = temp[2], src1.xyz = temp[3], src2.xyz = input[3]
>     MAD temp[7].xyz, -src0.xyz, src1.xxx, src2.xyz
> 25: src0.xyz = temp[0], src1.xyz = const[11]
>     MAD temp[14].xy, src0.xy_, src1.xy_, src0.000
> 26: TEX temp[2].xyz, temp[14].xy__, 2D[2];
> 27: src0.xyz = temp[2], src1.xyz = const[4]
>     MAD temp[2].xyz, src0.xyz, src1.xyz, src0.000
> 28: TEX temp[3].xyz, temp[7].xyz_, CUBE[3];
> 29: src0.xyz = temp[3], src1.xyz = temp[2]
>     MAD temp[5].xyz, src0.xyz, src0.111, -src1.xyz
> 30: src0.xyz = temp[5], src1.xyz = const[8], src2.xyz = temp[2]
>     MAD temp[5].xyz, src0.xyz, src1.xxx, src2.xyz
> 31: src0.xyz = temp[10]
>     MAD color[0].w, src0.x, src0.1, src0.y
> 32: src0.xyz = const[5], src1.xyz = temp[5]
>     MAD temp[13].xyz, src0.xyz, src0.111, -src1.xyz
> 33: src0.xyz = input[1], src1.xyz = temp[13], src2.xyz = temp[5]
>     MAD color[0].xyz, src0.xxx, src1.xyz, src2.xyz
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[2].yw, input[2].xy__, 2D[1];
>  2: TEX temp[1].yw, input[2].xy__, 2D[0];
>  3: src0.xyz = temp[2], src0.w = temp[2], src1.xyz = temp[1], src1.w = temp[1]
>     MAD temp[12].y, src0._y_, src0.111, -src1._y_
>     MAD temp[12].w, src0.w, src0.1, -src1.w
>  4: src0.xyz = const[0], src0.w = const[0], src1.xyz = temp[12], src1.w = temp[12], src2.xyz = temp[1], src2.w = temp[1]
>     MAD temp[3].y, src0._y_, src1._y_, src2._y_
>     MAD temp[3].w, src0.w, src1.w, src2.w
>  5: src0.xyz = const[7], src0.w = temp[3], src1.w = input[3]
>     MAD temp[3].x, src0.w__, src0.111, src0.000
>     MAD_SAT temp[1].w, src1.w, src0.z, src0.0
>  6: src0.xyz = temp[3], src1.xyz = const[7], src2.xyz = temp[10]
>     MAD temp[3].xy, src0.xy_, src1.xx_, -src0.11_
>     MAD color[0].w, src2.x, src0.1, src2.y
>  7: src0.xyz = temp[3], src0.w = input[0], src1.xyz = const[2]
>     MAD temp[3].xy, src0.xy_, src1.zz_, src0.000
>     RCP temp[11].w, src0.w
>  8: src0.xyz = input[0], src0.w = temp[11]
>     MAD temp[11].xy, src0.xy_, src0.ww_, src0.000
>  9: src0.xyz = temp[11], src1.xyz = const[9], src2.xyz = const[10]
>     MAD temp[11].xy, src0.xy_, src1.xy_, src2.xy_
> 10: src0.xyz = const[6], src1.xyz = temp[11]
>     MAD temp[0].y, src0._y_, src0.111, -src1._y_
> 11: src0.xyz = temp[11]
>     MAD temp[0].x, src0.x__, src0.111, src0.000
> 12: src0.xyz = temp[0], src1.xyz = const[11]
>     MAD temp[14].xy, src0.xy_, src1.xy_, src0.000
> 13: src0.xyz = temp[3], src0.w = temp[1]
>     MAD temp[3].xy, src0.xy_, src0.ww_, -src0.xy_
> 14: 
>     MAD temp[3].z, src0.__0, src0.111, src0.000
> 15: src0.xyz = temp[3]
>     DP3 temp[4].x, src0.xyz, src0.xyz
> 16: src0.xyz = temp[4]
>     MAD temp[4].x, src0.1__, src0.111, -src0.x__
> 17: src0.xyz = temp[4]
>     REPL_ALPHA temp[4].x
>     RSQ, |src0.x|
> 18: src0.xyz = temp[4]
>     REPL_ALPHA temp[3].z
>     RCP, src0.x
> 19: src0.xyz = temp[3], src1.xyz = input[4]
>     DP3 temp[2].x, src0.xyz, src1.xyz
> 20: src0.xyz = temp[3], src1.xyz = input[5]
>     DP3 temp[2].y, src0.xyz, src1.xyz
> 21: src0.xyz = temp[3], src1.xyz = input[6]
>     DP3 temp[2].z, src0.xyz, src1.xyz
> 22: src0.xyz = input[3], src1.xyz = temp[2]
>     DP3 temp[3].x, src0.xyz, src1.xyz
> 23: src0.xyz = temp[3], src1.xyz = const[7]
>     MAD temp[3].x, src0.x__, src1.x__, src0.000
> 24: src0.xyz = temp[2], src1.xyz = temp[3], src2.xyz = input[3]
>     MAD temp[7].xyz, -src0.xyz, src1.xxx, src2.xyz
> 25: BEGIN_TEX;
> 26: TEX temp[3].xyz, temp[7].xyz_, CUBE[3];
> 27: TEX temp[2].xyz, temp[14].xy__, 2D[2];
> 28: src0.xyz = temp[2], src1.xyz = const[4]
>     MAD temp[2].xyz, src0.xyz, src1.xyz, src0.000
> 29: src0.xyz = temp[3], src1.xyz = temp[2]
>     MAD temp[5].xyz, src0.xyz, src0.111, -src1.xyz
> 30: src0.xyz = temp[5], src1.xyz = const[8], src2.xyz = temp[2]
>     MAD temp[5].xyz, src0.xyz, src1.xxx, src2.xyz
> 31: src0.xyz = const[5], src1.xyz = temp[5]
>     MAD temp[13].xyz, src0.xyz, src0.111, -src1.xyz
> 32: src0.xyz = input[1], src1.xyz = temp[13], src2.xyz = temp[5]
>     MAD color[0].xyz, src0.xxx, src1.xyz, src2.xyz
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[8].yw, temp[1].xy__, 2D[1];
>  2: TEX temp[7].yw, temp[1].xy__, 2D[0];
>  3: src0.xyz = temp[8], src0.w = temp[8], src1.xyz = temp[7], src1.w = temp[7]
>     MAD temp[1].y, src0._y_, src0.111, -src1._y_
>     MAD temp[1].w, src0.w, src0.1, -src1.w
>  4: src0.xyz = const[0], src0.w = const[0], src1.xyz = temp[1], src1.w = temp[1], src2.xyz = temp[7], src2.w = temp[7]
>     MAD temp[9].y, src0._y_, src1._y_, src2._y_
>     MAD temp[9].w, src0.w, src1.w, src2.w
>  5: src0.xyz = const[7], src0.w = temp[9], src1.w = temp[2]
>     MAD temp[9].x, src0.w__, src0.111, src0.000
>     MAD_SAT temp[7].w, src1.w, src0.z, src0.0
>  6: src0.xyz = temp[9], src1.xyz = const[7], src2.xyz = temp[1]
>     MAD temp[9].xy, src0.xy_, src1.xx_, -src0.11_
>     MAD color[0].w, src2.x, src0.1, src2.y
>  7: src0.xyz = temp[9], src0.w = temp[6], src1.xyz = const[2]
>     MAD temp[9].xy, src0.xy_, src1.zz_, src0.000
>     RCP temp[10].w, src0.w
>  8: src0.xyz = temp[6], src0.w = temp[10]
>     MAD temp[10].xy, src0.xy_, src0.ww_, src0.000
>  9: src0.xyz = temp[10], src1.xyz = const[9], src2.xyz = const[10]
>     MAD temp[10].xy, src0.xy_, src1.xy_, src2.xy_
> 10: src0.xyz = const[6], src1.xyz = temp[10]
>     MAD temp[1].y, src0._y_, src0.111, -src1._y_
> 11: src0.xyz = temp[10]
>     MAD temp[1].x, src0.x__, src0.111, src0.000
> 12: src0.xyz = temp[1], src1.xyz = const[11]
>     MAD temp[6].xy, src0.xy_, src1.xy_, src0.000
> 13: src0.xyz = temp[9], src0.w = temp[7]
>     MAD temp[9].xy, src0.xy_, src0.ww_, -src0.xy_
> 14: 
>     MAD temp[9].z, src0.__0, src0.111, src0.000
> 15: src0.xyz = temp[9]
>     DP3 temp[1].x, src0.xyz, src0.xyz
> 16: src0.xyz = temp[1]
>     MAD temp[1].x, src0.1__, src0.111, -src0.x__
> 17: src0.xyz = temp[1]
>     REPL_ALPHA temp[1].x
>     RSQ, |src0.x|
> 18: src0.xyz = temp[1]
>     REPL_ALPHA temp[9].z
>     RCP, src0.x
> 19: src0.xyz = temp[9], src1.xyz = temp[3]
>     DP3 temp[8].x, src0.xyz, src1.xyz
> 20: src0.xyz = temp[9], src1.xyz = temp[4]
>     DP3 temp[8].y, src0.xyz, src1.xyz
> 21: src0.xyz = temp[9], src1.xyz = temp[5]
>     DP3 temp[8].z, src0.xyz, src1.xyz
> 22: src0.xyz = temp[2], src1.xyz = temp[8]
>     DP3 temp[9].x, src0.xyz, src1.xyz
> 23: src0.xyz = temp[9], src1.xyz = const[7]
>     MAD temp[9].x, src0.x__, src1.x__, src0.000
> 24: src0.xyz = temp[8], src1.xyz = temp[9], src2.xyz = temp[2]
>     MAD temp[1].xyz, -src0.xyz, src1.xxx, src2.xyz
> 25: BEGIN_TEX;
> 26: TEX temp[9].xyz, temp[1].xyz_, CUBE[3];
> 27: TEX temp[8].xyz, temp[6].xy__, 2D[2];
> 28: src0.xyz = temp[8], src1.xyz = const[4]
>     MAD temp[8].xyz, src0.xyz, src1.xyz, src0.000
> 29: src0.xyz = temp[9], src1.xyz = temp[8]
>     MAD temp[1].xyz, src0.xyz, src0.111, -src1.xyz
> 30: src0.xyz = temp[1], src1.xyz = const[8], src2.xyz = temp[8]
>     MAD temp[1].xyz, src0.xyz, src1.xxx, src2.xyz
> 31: src0.xyz = const[5], src1.xyz = temp[1]
>     MAD temp[2].xyz, src0.xyz, src0.111, -src1.xyz
> 32: src0.xyz = temp[0], src1.xyz = temp[2], src2.xyz = temp[1]
>     MAD color[0].xyz, src0.xxx, src1.xyz, src2.xyz
>pc=40*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 21, tex_end: 1  (code_addr: 00420540)
>  TEX:
>    TEX t8, t1, texture[1]   (00008a01)
>    TEX t7, t1, texture[0]   (000081c1)
>  0: xyz:  t8  t7  t0 -> t1.y                 (010401c8)
>       w:  t8  t7  t0 -> t1.w                 (008401c8)
>     xyz:   t8.xyz      1.0  -t7.xyz    op: 00090a80
>       w:     t8.w      1.0    -t7.w    op: 000a8889
>  1: xyz:  c0  t1  t7 -> t9.y                 (01247060)
>       w:  c0  t1  t7 -> t9.w                 (00a47060)
>     xyz:   c0.xyz   t1.xyz   t7.xyz    op: 00020200
>       w:     c0.w     t1.w     t7.w    op: 0002c509
>  2: xyz:  c7  t0  t0 -> t9.x                 (00a40027)
>       w:  t9  t2  t0 -> t7.w                 (009c0089)
>     xyz:   t9.www      1.0      0.0    op: 00050a8c
>       w:     t2.w     c7.z      0.0    op: 4004010a
>  3: xyz:  t9  c7  t1 -> t9.xy                (01a419c9)
>       w:  t0  t0  t0 -> o0.w                 (01000000)
>     xyz:   t9.xyz   c7.xxx     -1.0    op: 000d4280
>       w:     t1.x      1.0     t1.y    op: 0001c886
>  4: xyz:  t9  c2  t0 -> t9.xy                (01a40889)
>       w:  t6  t0  t0 -> t10.w                (00a80006)
>     xyz:   t9.xyz   c2.zzz      0.0    op: 00050380
>       w:     t6.w     t9.x     t9.x    op: 05000009
>  5: xyz:  t6  t0  t0 -> t10.xy               (01a80006)
>       w: t10  t0  t0 ->                      (0000000a)
>     xyz:   t6.xyz  t10.www      0.0    op: 00050600
>       w:     t6.x     t6.x     t6.x    op: 00000000
>  6: xyz: t10  c9 c10 -> t10.xy               (01aaaa4a)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:  t10.xyz   c9.xyz  c10.xyz    op: 00020200
>       w:    t10.x    t10.x    t10.x    op: 00000000
>  7: xyz:  c6 t10  t0 -> t1.y                 (010402a6)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   c6.xyz      1.0 -t10.xyz    op: 00090a80
>       w:     c6.x     c6.x     c6.x    op: 00000000
>  8: xyz: t10  t0  t0 -> t1.x                 (0084000a)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:  t10.xyz      1.0      0.0    op: 00050a80
>       w:    t10.x    t10.x    t10.x    op: 00000000
>  9: xyz:  t1 c11  t0 -> t6.xy                (01980ac1)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz  c11.xyz      0.0    op: 00050200
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 10: xyz:  t9  t0  t0 -> t9.xy                (01a40009)
>       w:  t7  t0  t0 ->                      (00000007)
>     xyz:   t9.xyz   t7.www  -t9.xyz    op: 00080600
>       w:     t9.x     t9.x     t9.x    op: 00000000
> 11: xyz:  t0  t0  t0 -> t9.z                 (02240000)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:      0.0      1.0      0.0    op: 00050a94
>       w:     t0.x     t0.x     t0.x    op: 00000000
> 12: xyz:  t9  t0  t0 -> t1.x                 (00840009)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t9.xyz   t9.xyz   t9.xxx    op: 00804000
>       w:     t9.x     t9.x     t9.x    op: 00000000
> 13: xyz:  t1  t0  t0 -> t1.x                 (00840001)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:      1.0      1.0  -t1.xyz    op: 00080a95
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 14: xyz:  t1  t0  t0 -> t1.x                 (00840001)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xxx   t1.xxx   t1.xxx    op: 05004081
>       w:   |t1.x|     t1.x     t1.x    op: 05800040
> 15: xyz:  t1  t0  t0 -> t9.z                 (02240001)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xxx   t1.xxx   t1.xxx    op: 05004081
>       w:     t1.x     t1.x     t1.x    op: 05000000
> 16: xyz:  t9  t3  t0 -> t8.x                 (00a000c9)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t9.xyz   t3.xyz   t9.xxx    op: 00804200
>       w:     t9.x     t9.x     t9.x    op: 00000000
> 17: xyz:  t9  t4  t0 -> t8.y                 (01200109)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t9.xyz   t4.xyz   t9.xxx    op: 00804200
>       w:     t9.x     t9.x     t9.x    op: 00000000
> 18: xyz:  t9  t5  t0 -> t8.z                 (02200149)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t9.xyz   t5.xyz   t9.xxx    op: 00804200
>       w:     t9.x     t9.x     t9.x    op: 00000000
> 19: xyz:  t2  t8  t0 -> t9.x                 (00a40202)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz   t8.xyz   t2.xxx    op: 00804200
>       w:     t2.x     t2.x     t2.x    op: 00000000
> 20: xyz:  t9  c7  t0 -> t9.x                 (00a409c9)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t9.xyz   c7.xyz      0.0    op: 00050200
>       w:     t9.x     t9.x     t9.x    op: 00000000
> 21: xyz:  t8  t9  t2 -> t1.xyz               (03842248)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:  -t8.xyz   t9.xxx   t2.xyz    op: 000202a0
>       w:     t8.x     t8.x     t8.x    op: 00000000
>NODE 1: alu_offset: 22, tex_offset: 2, alu_end: 4, tex_end: 1  (code_addr: 00422116)
>  TEX:
>    TEX t9, t1, texture[3]   (00009a41)
>    TEX t8, t6, texture[2]   (00009206)
> 22: xyz:  t8  c4  t0 -> t8.xyz               (03a00908)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t8.xyz   c4.xyz      0.0    op: 00050200
>       w:     t8.x     t8.x     t8.x    op: 00000000
> 23: xyz:  t9  t8  t0 -> t1.xyz               (03840209)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t9.xyz      1.0  -t8.xyz    op: 00090a80
>       w:     t9.x     t9.x     t9.x    op: 00000000
> 24: xyz:  t1  c8  t8 -> t1.xyz               (03848a01)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   c8.xxx   t8.xyz    op: 00020280
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 25: xyz:  c5  t1  t0 -> t2.xyz               (03880065)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   c5.xyz      1.0  -t1.xyz    op: 00090a80
>       w:     c5.x     c5.x     c5.x    op: 00000000
> 26: xyz:  t0  t2  t1 -> o0.xyz               (1c001080)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xxx   t2.xyz   t1.xyz    op: 00020201
>       w:     t0.x     t0.x     t0.x    op: 00000000
>r300: Initial fragment program
>FRAG
>DCL IN[0], POSITION, LINEAR
>DCL IN[1], COLOR[1], LINEAR
>DCL IN[2], GENERIC[0], PERSPECTIVE
>DCL IN[3], GENERIC[2], PERSPECTIVE
>DCL IN[4], GENERIC[3], PERSPECTIVE
>DCL IN[5], GENERIC[4], PERSPECTIVE
>DCL IN[6], GENERIC[5], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL SAMP[1]
>DCL SAMP[2]
>DCL SAMP[3]
>DCL CONST[5..6]
>DCL CONST[0]
>DCL CONST[2]
>DCL CONST[4]
>DCL TEMP[0..10]
>IMM FLT32 {    2.0000,     1.0000,     0.0001,     0.0000}
>IMM FLT32 {    0.1250,     0.0000,     0.0000,     0.0000}
>  0: MOV TEMP[0], IN[0]
>  1: SUB TEMP[0].y, CONST[6], IN[0]
>  2: TEX TEMP[1], IN[2], SAMP[0], 2D
>  3: TEX TEMP[2], IN[2], SAMP[1], 2D
>  4: LRP TEMP[3], CONST[0], TEMP[2], TEMP[1]
>  5: MOV TEMP[3].x, TEMP[3].wwww
>  6: MAD TEMP[3], TEMP[3], IMM[0].xxxx, -IMM[0].yyyy
>  7: MUL TEMP[3].xy, TEMP[3], CONST[2].zzzz
>  8: MUL_SAT TEMP[1].w, IN[3].wwww, IMM[0].zzzz
>  9: SUB TEMP[2], IMM[0].wwww, TEMP[3]
> 10: MAD TEMP[3], TEMP[3], TEMP[1].wwww, TEMP[2]
> 11: MOV TEMP[3].z, IMM[0].wwww
> 12: DP3 TEMP[4].x, TEMP[3], TEMP[3]
> 13: ADD TEMP[4].x, IMM[0].yyyy, -TEMP[4].xxxx
> 14: RSQ TEMP[4].x, TEMP[4].xxxx
> 15: RCP TEMP[3].z, TEMP[4].xxxx
> 16: DP3 TEMP[2].x, TEMP[3], IN[4]
> 17: DP3 TEMP[2].y, TEMP[3], IN[5]
> 18: DP3 TEMP[2].z, TEMP[3], IN[6]
> 19: MOV TEMP[5], TEMP[0]
> 20: MOV TEMP[6], IN[3]
> 21: DP3 TEMP[3], TEMP[6], TEMP[2]
> 22: MUL TEMP[3], TEMP[3], IMM[0].xxxx
> 23: MAD TEMP[7], -TEMP[2], TEMP[3].xxxx, TEMP[6]
> 24: MOV TEMP[8], IMM[1].xxxx
> 25: TEX TEMP[2], TEMP[5], SAMP[2], RECT
> 26: MUL TEMP[2], TEMP[2], CONST[4]
> 27: TEX TEMP[3], TEMP[7], SAMP[3], CUBE
> 28: SUB TEMP[5], TEMP[3], TEMP[2]
> 29: MAD TEMP[5], TEMP[5], TEMP[8], TEMP[2]
> 30: MOV TEMP[9], TEMP[5]
> 31: ADD OUT[0].w, TEMP[10].xxxx, TEMP[10].yyyy
> 32: LRP OUT[0].xyz, IN[1].xxxx, CONST[5], TEMP[9]
> 33: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: RCP temp[11].w, input[0].wwww;
>  1: MUL temp[11].xyz, input[0], temp[11].wwww;
>  2: MAD temp[11].xyz, temp[11].xyz0, const[9].xyz0, const[10].xyz0;
>  3: MOV temp[0], temp[11];
>  4: SUB temp[0].y, const[6], temp[11];
>  5: TEX temp[1], input[2], 2D[0];
>  6: TEX temp[2], input[2], 2D[1];
>  7: LRP temp[3], const[0], temp[2], temp[1];
>  8: MOV temp[3].x, temp[3].wwww;
>  9: MAD temp[3], temp[3], const[7].xxxx, -const[7].yyyy;
> 10: MUL temp[3].xy, temp[3], const[2].zzzz;
> 11: MUL_SAT temp[1].w, input[3].wwww, const[7].zzzz;
> 12: SUB temp[2], const[7].wwww, temp[3];
> 13: MAD temp[3], temp[3], temp[1].wwww, temp[2];
> 14: MOV temp[3].z, const[7].wwww;
> 15: DP3 temp[4].x, temp[3], temp[3];
> 16: ADD temp[4].x, const[7].yyyy, -temp[4].xxxx;
> 17: RSQ temp[4].x, temp[4].xxxx;
> 18: RCP temp[3].z, temp[4].xxxx;
> 19: DP3 temp[2].x, temp[3], input[4];
> 20: DP3 temp[2].y, temp[3], input[5];
> 21: DP3 temp[2].z, temp[3], input[6];
> 22: MOV temp[5], temp[0];
> 23: MOV temp[6], input[3];
> 24: DP3 temp[3], temp[6], temp[2];
> 25: MUL temp[3], temp[3], const[7].xxxx;
> 26: MAD temp[7], -temp[2], temp[3].xxxx, temp[6];
> 27: MOV temp[8], const[8].xxxx;
> 28: TEX temp[2], temp[5], RECT[2];
> 29: MUL temp[2], temp[2], const[4];
> 30: TEX temp[3], temp[7], CUBE[3];
> 31: SUB temp[5], temp[3], temp[2];
> 32: MAD temp[5], temp[5], temp[8], temp[2];
> 33: MOV temp[9], temp[5];
> 34: ADD output[0].w, temp[10].xxxx, temp[10].yyyy;
> 35: LRP output[0].xyz, input[1].xxxx, const[5], temp[9];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: RCP temp[11].w, input[0].wwww;
>  1: MUL temp[11].xyz, input[0], temp[11].wwww;
>  2: MAD temp[11].xyz, temp[11].xyz0, const[9].xyz0, const[10].xyz0;
>  3: MOV temp[0], temp[11];
>  4: SUB temp[0].y, const[6], temp[11];
>  5: TEX temp[1], input[2], 2D[0];
>  6: TEX temp[2], input[2], 2D[1];
>  7: LRP temp[3], const[0], temp[2], temp[1];
>  8: MOV temp[3].x, temp[3].wwww;
>  9: MAD temp[3], temp[3], const[7].xxxx, -const[7].yyyy;
> 10: MUL temp[3].xy, temp[3], const[2].zzzz;
> 11: MUL_SAT temp[1].w, input[3].wwww, const[7].zzzz;
> 12: SUB temp[2], const[7].wwww, temp[3];
> 13: MAD temp[3], temp[3], temp[1].wwww, temp[2];
> 14: MOV temp[3].z, const[7].wwww;
> 15: DP3 temp[4].x, temp[3], temp[3];
> 16: ADD temp[4].x, const[7].yyyy, -temp[4].xxxx;
> 17: RSQ temp[4].x, temp[4].xxxx;
> 18: RCP temp[3].z, temp[4].xxxx;
> 19: DP3 temp[2].x, temp[3], input[4];
> 20: DP3 temp[2].y, temp[3], input[5];
> 21: DP3 temp[2].z, temp[3], input[6];
> 22: MOV temp[5], temp[0];
> 23: MOV temp[6], input[3];
> 24: DP3 temp[3], temp[6], temp[2];
> 25: MUL temp[3], temp[3], const[7].xxxx;
> 26: MAD temp[7], -temp[2], temp[3].xxxx, temp[6];
> 27: MOV temp[8], const[8].xxxx;
> 28: TEX temp[2], temp[5], RECT[2];
> 29: MUL temp[2], temp[2], const[4];
> 30: TEX temp[3], temp[7], CUBE[3];
> 31: SUB temp[5], temp[3], temp[2];
> 32: MAD temp[5], temp[5], temp[8], temp[2];
> 33: MOV temp[9], temp[5];
> 34: ADD output[0].w, temp[10].xxxx, temp[10].yyyy;
> 35: LRP output[0].xyz, input[1].xxxx, const[5], temp[9];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: RCP temp[11].w, input[0].wwww;
>  1: MUL temp[11].xyz, input[0], temp[11].wwww;
>  2: MAD temp[11].xyz, temp[11].xyz0, const[9].xyz0, const[10].xyz0;
>  3: MOV temp[0], temp[11];
>  4: SUB temp[0].y, const[6], temp[11];
>  5: TEX temp[1], input[2], 2D[0];
>  6: TEX temp[2], input[2], 2D[1];
>  7: LRP temp[3], const[0], temp[2], temp[1];
>  8: MOV temp[3].x, temp[3].wwww;
>  9: MAD temp[3], temp[3], const[7].xxxx, -const[7].yyyy;
> 10: MUL temp[3].xy, temp[3], const[2].zzzz;
> 11: MUL_SAT temp[1].w, input[3].wwww, const[7].zzzz;
> 12: SUB temp[2], const[7].wwww, temp[3];
> 13: MAD temp[3], temp[3], temp[1].wwww, temp[2];
> 14: MOV temp[3].z, const[7].wwww;
> 15: DP3 temp[4].x, temp[3], temp[3];
> 16: ADD temp[4].x, const[7].yyyy, -temp[4].xxxx;
> 17: RSQ temp[4].x, temp[4].xxxx;
> 18: RCP temp[3].z, temp[4].xxxx;
> 19: DP3 temp[2].x, temp[3], input[4];
> 20: DP3 temp[2].y, temp[3], input[5];
> 21: DP3 temp[2].z, temp[3], input[6];
> 22: MOV temp[5], temp[0];
> 23: MOV temp[6], input[3];
> 24: DP3 temp[3], temp[6], temp[2];
> 25: MUL temp[3], temp[3], const[7].xxxx;
> 26: MAD temp[7], -temp[2], temp[3].xxxx, temp[6];
> 27: MOV temp[8], const[8].xxxx;
> 28: TEX temp[2], temp[5], RECT[2];
> 29: MUL temp[2], temp[2], const[4];
> 30: TEX temp[3], temp[7], CUBE[3];
> 31: SUB temp[5], temp[3], temp[2];
> 32: MAD temp[5], temp[5], temp[8], temp[2];
> 33: MOV temp[9], temp[5];
> 34: ADD output[0].w, temp[10].xxxx, temp[10].yyyy;
> 35: LRP output[0].xyz, input[1].xxxx, const[5], temp[9];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: RCP temp[11].w, input[0].wwww;
>  1: MUL temp[11].xyz, input[0], temp[11].wwww;
>  2: MAD temp[11].xyz, temp[11].xyz0, const[9].xyz0, const[10].xyz0;
>  3: MOV temp[0], temp[11];
>  4: ADD temp[0].y, const[6], -temp[11];
>  5: TEX temp[1], input[2], 2D[0];
>  6: TEX temp[2], input[2], 2D[1];
>  7: ADD temp[12], temp[2], -temp[1];
>  8: MAD temp[3], const[0], temp[12], temp[1];
>  9: MOV temp[3].x, temp[3].wwww;
> 10: MAD temp[3], temp[3], const[7].xxxx, -const[7].yyyy;
> 11: MUL temp[3].xy, temp[3], const[2].zzzz;
> 12: MUL_SAT temp[1].w, input[3].wwww, const[7].zzzz;
> 13: ADD temp[2], const[7].wwww, -temp[3];
> 14: MAD temp[3], temp[3], temp[1].wwww, temp[2];
> 15: MOV temp[3].z, const[7].wwww;
> 16: DP3 temp[4].x, temp[3], temp[3];
> 17: ADD temp[4].x, const[7].yyyy, -temp[4].xxxx;
> 18: RSQ temp[4].x, |temp[4].xxxx|;
> 19: RCP temp[3].z, temp[4].xxxx;
> 20: DP3 temp[2].x, temp[3], input[4];
> 21: DP3 temp[2].y, temp[3], input[5];
> 22: DP3 temp[2].z, temp[3], input[6];
> 23: MOV temp[5], temp[0];
> 24: MOV temp[6], input[3];
> 25: DP3 temp[3], temp[6], temp[2];
> 26: MUL temp[3], temp[3], const[7].xxxx;
> 27: MAD temp[7], -temp[2], temp[3].xxxx, temp[6];
> 28: MOV temp[8], const[8].xxxx;
> 29: TEX temp[2], temp[5], RECT[2];
> 30: MUL temp[2], temp[2], const[4];
> 31: TEX temp[3], temp[7], CUBE[3];
> 32: ADD temp[5], temp[3], -temp[2];
> 33: MAD temp[5], temp[5], temp[8], temp[2];
> 34: MOV temp[9], temp[5];
> 35: ADD output[0].w, temp[10].xxxx, temp[10].yyyy;
> 36: ADD temp[13], const[5], -temp[9];
> 37: MAD output[0].xyz, input[1].xxxx, temp[13], temp[9];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: RCP temp[11].w, input[0].wwww;
>  1: MUL temp[11].xyz, input[0], temp[11].wwww;
>  2: MAD temp[11].xyz, temp[11].xyz0, const[9].xyz0, const[10].xyz0;
>  3: MOV temp[0], temp[11];
>  4: ADD temp[0].y, const[6], -temp[11];
>  5: TEX temp[1], input[2], 2D[0];
>  6: TEX temp[2], input[2], 2D[1];
>  7: ADD temp[12], temp[2], -temp[1];
>  8: MAD temp[3], const[0], temp[12], temp[1];
>  9: MOV temp[3].x, temp[3].wwww;
> 10: MAD temp[3], temp[3], const[7].xxxx, -const[7].yyyy;
> 11: MUL temp[3].xy, temp[3], const[2].zzzz;
> 12: MUL_SAT temp[1].w, input[3].wwww, const[7].zzzz;
> 13: ADD temp[2], const[7].wwww, -temp[3];
> 14: MAD temp[3], temp[3], temp[1].wwww, temp[2];
> 15: MOV temp[3].z, const[7].wwww;
> 16: DP3 temp[4].x, temp[3], temp[3];
> 17: ADD temp[4].x, const[7].yyyy, -temp[4].xxxx;
> 18: RSQ temp[4].x, |temp[4].xxxx|;
> 19: RCP temp[3].z, temp[4].xxxx;
> 20: DP3 temp[2].x, temp[3], input[4];
> 21: DP3 temp[2].y, temp[3], input[5];
> 22: DP3 temp[2].z, temp[3], input[6];
> 23: MOV temp[5], temp[0];
> 24: MOV temp[6], input[3];
> 25: DP3 temp[3], temp[6], temp[2];
> 26: MUL temp[3], temp[3], const[7].xxxx;
> 27: MAD temp[7], -temp[2], temp[3].xxxx, temp[6];
> 28: MOV temp[8], const[8].xxxx;
> 29: MUL temp[14], temp[5], const[11];
> 30: TEX temp[2], temp[14], 2D[2];
> 31: MUL temp[2], temp[2], const[4];
> 32: TEX temp[3], temp[7], CUBE[3];
> 33: ADD temp[5], temp[3], -temp[2];
> 34: MAD temp[5], temp[5], temp[8], temp[2];
> 35: MOV temp[9], temp[5];
> 36: ADD output[0].w, temp[10].xxxx, temp[10].yyyy;
> 37: ADD temp[13], const[5], -temp[9];
> 38: MAD output[0].xyz, input[1].xxxx, temp[13], temp[9];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: RCP temp[11].w, input[0].w___;
>  1: MUL temp[11].xy, input[0].xy__, temp[11].ww__;
>  2: MAD temp[11].xy, temp[11].xy__, const[9].xy__, const[10].xy__;
>  3: MOV temp[0].x, temp[11].x___;
>  4: ADD temp[0].y, const[6]._y__, -temp[11]._y__;
>  5: TEX temp[1].yw, input[2].xy__, 2D[0];
>  6: TEX temp[2].yw, input[2].xy__, 2D[1];
>  7: ADD temp[12].yw, temp[2]._y_w, -temp[1]._y_w;
>  8: MAD temp[3].yw, const[0]._y_w, temp[12]._y_w, temp[1]._y_w;
>  9: MOV temp[3].x, temp[3].w___;
> 10: MAD temp[3].xy, temp[3].xy__, const[7].xx__, -const[7].yy__;
> 11: MUL temp[3].xy, temp[3].xy__, const[2].zz__;
> 12: MUL_SAT temp[1].w, input[3].___w, const[7].___z;
> 13: ADD temp[2].xy, const[7].ww__, -temp[3].xy__;
> 14: MAD temp[3].xy, temp[3].xy__, temp[1].ww__, temp[2].xy__;
> 15: MOV temp[3].z, const[7].__w_;
> 16: DP3 temp[4].x, temp[3].xyz_, temp[3].xyz_;
> 17: ADD temp[4].x, const[7].y___, -temp[4].x___;
> 18: RSQ temp[4].x, |temp[4].x___|;
> 19: RCP temp[3].z, temp[4].x___;
> 20: DP3 temp[2].x, temp[3].xyz_, input[4].xyz_;
> 21: DP3 temp[2].y, temp[3].xyz_, input[5].xyz_;
> 22: DP3 temp[2].z, temp[3].xyz_, input[6].xyz_;
> 23: MOV temp[5].xy, temp[0].xy__;
> 24: MOV temp[6].xyz, input[3].xyz_;
> 25: DP3 temp[3].x, temp[6].xyz_, temp[2].xyz_;
> 26: MUL temp[3].x, temp[3].x___, const[7].x___;
> 27: MAD temp[7].xyz, -temp[2].xyz_, temp[3].xxx_, temp[6].xyz_;
> 28: MOV temp[8].xyz, const[8].xxx_;
> 29: MUL temp[14].xy, temp[5].xy__, const[11].xy__;
> 30: TEX temp[2].xyz, temp[14].xy__, 2D[2];
> 31: MUL temp[2].xyz, temp[2].xyz_, const[4].xyz_;
> 32: TEX temp[3].xyz, temp[7].xyz_, CUBE[3];
> 33: ADD temp[5].xyz, temp[3].xyz_, -temp[2].xyz_;
> 34: MAD temp[5].xyz, temp[5].xyz_, temp[8].xyz_, temp[2].xyz_;
> 35: MOV temp[9].xyz, temp[5].xyz_;
> 36: ADD output[0].w, temp[10].___x, temp[10].___y;
> 37: ADD temp[13].xyz, const[5].xyz_, -temp[9].xyz_;
> 38: MAD output[0].xyz, input[1].xxx_, temp[13].xyz_, temp[9].xyz_;
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: RCP temp[11].w, input[0].w___;
>  1: MUL temp[11].xy, input[0].xy__, temp[11].ww__;
>  2: MAD temp[11].xy, temp[11].xy__, const[9].xy__, const[10].xy__;
>  3: MOV temp[0].x, temp[11].x___;
>  4: ADD temp[0].y, const[6]._y__, -temp[11]._y__;
>  5: TEX temp[1].yw, input[2].xy__, 2D[0];
>  6: TEX temp[2].yw, input[2].xy__, 2D[1];
>  7: ADD temp[12].yw, temp[2]._y_w, -temp[1]._y_w;
>  8: MAD temp[3].yw, const[0]._y_w, temp[12]._y_w, temp[1]._y_w;
>  9: MOV temp[3].x, temp[3].w___;
> 10: MAD temp[3].xy, temp[3].xy__, const[7].xx__, -none.11__;
> 11: MUL temp[3].xy, temp[3].xy__, const[2].zz__;
> 12: MUL_SAT temp[1].w, input[3].___w, const[7].___z;
> 13: MAD temp[3].xy, temp[3].xy__, temp[1].ww__, temp[3].-x-y__;
> 14: MOV temp[3].z, none.__0_;
> 15: DP3 temp[4].x, temp[3].xyz_, temp[3].xyz_;
> 16: ADD temp[4].x, none.1___, -temp[4].x___;
> 17: RSQ temp[4].x, |temp[4].x___|;
> 18: RCP temp[3].z, temp[4].x___;
> 19: DP3 temp[2].x, temp[3].xyz_, input[4].xyz_;
> 20: DP3 temp[2].y, temp[3].xyz_, input[5].xyz_;
> 21: DP3 temp[2].z, temp[3].xyz_, input[6].xyz_;
> 22: DP3 temp[3].x, input[3].xyz_, temp[2].xyz_;
> 23: MUL temp[3].x, temp[3].x___, const[7].x___;
> 24: MAD temp[7].xyz, -temp[2].xyz_, temp[3].xxx_, input[3].xyz_;
> 25: MUL temp[14].xy, temp[0].xy__, const[11].xy__;
> 26: TEX temp[2].xyz, temp[14].xy__, 2D[2];
> 27: MUL temp[2].xyz, temp[2].xyz_, const[4].xyz_;
> 28: TEX temp[3].xyz, temp[7].xyz_, CUBE[3];
> 29: ADD temp[5].xyz, temp[3].xyz_, -temp[2].xyz_;
> 30: MAD temp[5].xyz, temp[5].xyz_, const[8].xxx_, temp[2].xyz_;
> 31: ADD output[0].w, temp[10].___x, temp[10].___y;
> 32: ADD temp[13].xyz, const[5].xyz_, -temp[5].xyz_;
> 33: MAD output[0].xyz, input[1].xxx_, temp[13].xyz_, temp[5].xyz_;
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: RCP temp[11].w, input[0].w___;
>  1: MUL temp[11].xy, input[0].xy__, temp[11].ww__;
>  2: MAD temp[11].xy, temp[11].xy__, const[9].xy__, const[10].xy__;
>  3: MOV temp[0].x, temp[11].x___;
>  4: ADD temp[0].y, const[6]._y__, -temp[11]._y__;
>  5: TEX temp[1].yw, input[2].xy__, 2D[0];
>  6: TEX temp[2].yw, input[2].xy__, 2D[1];
>  7: ADD temp[12].yw, temp[2]._y_w, -temp[1]._y_w;
>  8: MAD temp[3].yw, const[0]._y_w, temp[12]._y_w, temp[1]._y_w;
>  9: MOV temp[3].x, temp[3].w___;
> 10: MAD temp[3].xy, temp[3].xy__, const[7].xx__, -none.11__;
> 11: MUL temp[3].xy, temp[3].xy__, const[2].zz__;
> 12: MUL_SAT temp[1].w, input[3].___w, const[7].___z;
> 13: MAD temp[3].xy, temp[3].xy__, temp[1].ww__, temp[3].-x-y__;
> 14: MOV temp[3].z, none.__0_;
> 15: DP3 temp[4].x, temp[3].xyz_, temp[3].xyz_;
> 16: ADD temp[4].x, none.1___, -temp[4].x___;
> 17: RSQ temp[4].x, |temp[4].x___|;
> 18: RCP temp[3].z, temp[4].x___;
> 19: DP3 temp[2].x, temp[3].xyz_, input[4].xyz_;
> 20: DP3 temp[2].y, temp[3].xyz_, input[5].xyz_;
> 21: DP3 temp[2].z, temp[3].xyz_, input[6].xyz_;
> 22: DP3 temp[3].x, input[3].xyz_, temp[2].xyz_;
> 23: MUL temp[3].x, temp[3].x___, const[7].x___;
> 24: MAD temp[7].xyz, -temp[2].xyz_, temp[3].xxx_, input[3].xyz_;
> 25: MUL temp[14].xy, temp[0].xy__, const[11].xy__;
> 26: TEX temp[2].xyz, temp[14].xy__, 2D[2];
> 27: MUL temp[2].xyz, temp[2].xyz_, const[4].xyz_;
> 28: TEX temp[3].xyz, temp[7].xyz_, CUBE[3];
> 29: ADD temp[5].xyz, temp[3].xyz_, -temp[2].xyz_;
> 30: MAD temp[5].xyz, temp[5].xyz_, const[8].xxx_, temp[2].xyz_;
> 31: ADD output[0].w, temp[10].___x, temp[10].___y;
> 32: ADD temp[13].xyz, const[5].xyz_, -temp[5].xyz_;
> 33: MAD output[0].xyz, input[1].xxx_, temp[13].xyz_, temp[5].xyz_;
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: src0.w = input[0]
>     RCP temp[11].w, src0.w
>  1: src0.xyz = input[0], src0.w = temp[11]
>     MAD temp[11].xy, src0.xy_, src0.ww_, src0.000
>  2: src0.xyz = temp[11], src1.xyz = const[9], src2.xyz = const[10]
>     MAD temp[11].xy, src0.xy_, src1.xy_, src2.xy_
>  3: src0.xyz = temp[11]
>     MAD temp[0].x, src0.x__, src0.111, src0.000
>  4: src0.xyz = const[6], src1.xyz = temp[11]
>     MAD temp[0].y, src0._y_, src0.111, -src1._y_
>  5: TEX temp[1].yw, input[2].xy__, 2D[0];
>  6: TEX temp[2].yw, input[2].xy__, 2D[1];
>  7: src0.xyz = temp[2], src0.w = temp[2], src1.xyz = temp[1], src1.w = temp[1]
>     MAD temp[12].y, src0._y_, src0.111, -src1._y_
>     MAD temp[12].w, src0.w, src0.1, -src1.w
>  8: src0.xyz = const[0], src0.w = const[0], src1.xyz = temp[12], src1.w = temp[12], src2.xyz = temp[1], src2.w = temp[1]
>     MAD temp[3].y, src0._y_, src1._y_, src2._y_
>     MAD temp[3].w, src0.w, src1.w, src2.w
>  9: src0.w = temp[3]
>     MAD temp[3].x, src0.w__, src0.111, src0.000
> 10: src0.xyz = temp[3], src1.xyz = const[7]
>     MAD temp[3].xy, src0.xy_, src1.xx_, -src0.11_
> 11: src0.xyz = temp[3], src1.xyz = const[2]
>     MAD temp[3].xy, src0.xy_, src1.zz_, src0.000
> 12: src0.xyz = const[7], src0.w = input[3]
>     MAD_SAT temp[1].w, src0.w, src0.z, src0.0
> 13: src0.xyz = temp[3], src0.w = temp[1]
>     MAD temp[3].xy, src0.xy_, src0.ww_, -src0.xy_
> 14: 
>     MAD temp[3].z, src0.__0, src0.111, src0.000
> 15: src0.xyz = temp[3]
>     DP3 temp[4].x, src0.xyz, src0.xyz
> 16: src0.xyz = temp[4]
>     MAD temp[4].x, src0.1__, src0.111, -src0.x__
> 17: src0.xyz = temp[4]
>     REPL_ALPHA temp[4].x
>     RSQ, |src0.x|
> 18: src0.xyz = temp[4]
>     REPL_ALPHA temp[3].z
>     RCP, src0.x
> 19: src0.xyz = temp[3], src1.xyz = input[4]
>     DP3 temp[2].x, src0.xyz, src1.xyz
> 20: src0.xyz = temp[3], src1.xyz = input[5]
>     DP3 temp[2].y, src0.xyz, src1.xyz
> 21: src0.xyz = temp[3], src1.xyz = input[6]
>     DP3 temp[2].z, src0.xyz, src1.xyz
> 22: src0.xyz = input[3], src1.xyz = temp[2]
>     DP3 temp[3].x, src0.xyz, src1.xyz
> 23: src0.xyz = temp[3], src1.xyz = const[7]
>     MAD temp[3].x, src0.x__, src1.x__, src0.000
> 24: src0.xyz = temp[2], src1.xyz = temp[3], src2.xyz = input[3]
>     MAD temp[7].xyz, -src0.xyz, src1.xxx, src2.xyz
> 25: src0.xyz = temp[0], src1.xyz = const[11]
>     MAD temp[14].xy, src0.xy_, src1.xy_, src0.000
> 26: TEX temp[2].xyz, temp[14].xy__, 2D[2];
> 27: src0.xyz = temp[2], src1.xyz = const[4]
>     MAD temp[2].xyz, src0.xyz, src1.xyz, src0.000
> 28: TEX temp[3].xyz, temp[7].xyz_, CUBE[3];
> 29: src0.xyz = temp[3], src1.xyz = temp[2]
>     MAD temp[5].xyz, src0.xyz, src0.111, -src1.xyz
> 30: src0.xyz = temp[5], src1.xyz = const[8], src2.xyz = temp[2]
>     MAD temp[5].xyz, src0.xyz, src1.xxx, src2.xyz
> 31: src0.xyz = temp[10]
>     MAD color[0].w, src0.x, src0.1, src0.y
> 32: src0.xyz = const[5], src1.xyz = temp[5]
>     MAD temp[13].xyz, src0.xyz, src0.111, -src1.xyz
> 33: src0.xyz = input[1], src1.xyz = temp[13], src2.xyz = temp[5]
>     MAD color[0].xyz, src0.xxx, src1.xyz, src2.xyz
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[2].yw, input[2].xy__, 2D[1];
>  2: TEX temp[1].yw, input[2].xy__, 2D[0];
>  3: src0.xyz = temp[2], src0.w = temp[2], src1.xyz = temp[1], src1.w = temp[1]
>     MAD temp[12].y, src0._y_, src0.111, -src1._y_
>     MAD temp[12].w, src0.w, src0.1, -src1.w
>  4: src0.xyz = const[0], src0.w = const[0], src1.xyz = temp[12], src1.w = temp[12], src2.xyz = temp[1], src2.w = temp[1]
>     MAD temp[3].y, src0._y_, src1._y_, src2._y_
>     MAD temp[3].w, src0.w, src1.w, src2.w
>  5: src0.xyz = const[7], src0.w = temp[3], src1.w = input[3]
>     MAD temp[3].x, src0.w__, src0.111, src0.000
>     MAD_SAT temp[1].w, src1.w, src0.z, src0.0
>  6: src0.xyz = temp[3], src1.xyz = const[7], src2.xyz = temp[10]
>     MAD temp[3].xy, src0.xy_, src1.xx_, -src0.11_
>     MAD color[0].w, src2.x, src0.1, src2.y
>  7: src0.xyz = temp[3], src0.w = input[0], src1.xyz = const[2]
>     MAD temp[3].xy, src0.xy_, src1.zz_, src0.000
>     RCP temp[11].w, src0.w
>  8: src0.xyz = input[0], src0.w = temp[11]
>     MAD temp[11].xy, src0.xy_, src0.ww_, src0.000
>  9: src0.xyz = temp[11], src1.xyz = const[9], src2.xyz = const[10]
>     MAD temp[11].xy, src0.xy_, src1.xy_, src2.xy_
> 10: src0.xyz = const[6], src1.xyz = temp[11]
>     MAD temp[0].y, src0._y_, src0.111, -src1._y_
> 11: src0.xyz = temp[11]
>     MAD temp[0].x, src0.x__, src0.111, src0.000
> 12: src0.xyz = temp[0], src1.xyz = const[11]
>     MAD temp[14].xy, src0.xy_, src1.xy_, src0.000
> 13: src0.xyz = temp[3], src0.w = temp[1]
>     MAD temp[3].xy, src0.xy_, src0.ww_, -src0.xy_
> 14: 
>     MAD temp[3].z, src0.__0, src0.111, src0.000
> 15: src0.xyz = temp[3]
>     DP3 temp[4].x, src0.xyz, src0.xyz
> 16: src0.xyz = temp[4]
>     MAD temp[4].x, src0.1__, src0.111, -src0.x__
> 17: src0.xyz = temp[4]
>     REPL_ALPHA temp[4].x
>     RSQ, |src0.x|
> 18: src0.xyz = temp[4]
>     REPL_ALPHA temp[3].z
>     RCP, src0.x
> 19: src0.xyz = temp[3], src1.xyz = input[4]
>     DP3 temp[2].x, src0.xyz, src1.xyz
> 20: src0.xyz = temp[3], src1.xyz = input[5]
>     DP3 temp[2].y, src0.xyz, src1.xyz
> 21: src0.xyz = temp[3], src1.xyz = input[6]
>     DP3 temp[2].z, src0.xyz, src1.xyz
> 22: src0.xyz = input[3], src1.xyz = temp[2]
>     DP3 temp[3].x, src0.xyz, src1.xyz
> 23: src0.xyz = temp[3], src1.xyz = const[7]
>     MAD temp[3].x, src0.x__, src1.x__, src0.000
> 24: src0.xyz = temp[2], src1.xyz = temp[3], src2.xyz = input[3]
>     MAD temp[7].xyz, -src0.xyz, src1.xxx, src2.xyz
> 25: BEGIN_TEX;
> 26: TEX temp[3].xyz, temp[7].xyz_, CUBE[3];
> 27: TEX temp[2].xyz, temp[14].xy__, 2D[2];
> 28: src0.xyz = temp[2], src1.xyz = const[4]
>     MAD temp[2].xyz, src0.xyz, src1.xyz, src0.000
> 29: src0.xyz = temp[3], src1.xyz = temp[2]
>     MAD temp[5].xyz, src0.xyz, src0.111, -src1.xyz
> 30: src0.xyz = temp[5], src1.xyz = const[8], src2.xyz = temp[2]
>     MAD temp[5].xyz, src0.xyz, src1.xxx, src2.xyz
> 31: src0.xyz = const[5], src1.xyz = temp[5]
>     MAD temp[13].xyz, src0.xyz, src0.111, -src1.xyz
> 32: src0.xyz = input[1], src1.xyz = temp[13], src2.xyz = temp[5]
>     MAD color[0].xyz, src0.xxx, src1.xyz, src2.xyz
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[8].yw, temp[1].xy__, 2D[1];
>  2: TEX temp[7].yw, temp[1].xy__, 2D[0];
>  3: src0.xyz = temp[8], src0.w = temp[8], src1.xyz = temp[7], src1.w = temp[7]
>     MAD temp[1].y, src0._y_, src0.111, -src1._y_
>     MAD temp[1].w, src0.w, src0.1, -src1.w
>  4: src0.xyz = const[0], src0.w = const[0], src1.xyz = temp[1], src1.w = temp[1], src2.xyz = temp[7], src2.w = temp[7]
>     MAD temp[9].y, src0._y_, src1._y_, src2._y_
>     MAD temp[9].w, src0.w, src1.w, src2.w
>  5: src0.xyz = const[7], src0.w = temp[9], src1.w = temp[2]
>     MAD temp[9].x, src0.w__, src0.111, src0.000
>     MAD_SAT temp[7].w, src1.w, src0.z, src0.0
>  6: src0.xyz = temp[9], src1.xyz = const[7], src2.xyz = temp[1]
>     MAD temp[9].xy, src0.xy_, src1.xx_, -src0.11_
>     MAD color[0].w, src2.x, src0.1, src2.y
>  7: src0.xyz = temp[9], src0.w = temp[6], src1.xyz = const[2]
>     MAD temp[9].xy, src0.xy_, src1.zz_, src0.000
>     RCP temp[10].w, src0.w
>  8: src0.xyz = temp[6], src0.w = temp[10]
>     MAD temp[10].xy, src0.xy_, src0.ww_, src0.000
>  9: src0.xyz = temp[10], src1.xyz = const[9], src2.xyz = const[10]
>     MAD temp[10].xy, src0.xy_, src1.xy_, src2.xy_
> 10: src0.xyz = const[6], src1.xyz = temp[10]
>     MAD temp[1].y, src0._y_, src0.111, -src1._y_
> 11: src0.xyz = temp[10]
>     MAD temp[1].x, src0.x__, src0.111, src0.000
> 12: src0.xyz = temp[1], src1.xyz = const[11]
>     MAD temp[6].xy, src0.xy_, src1.xy_, src0.000
> 13: src0.xyz = temp[9], src0.w = temp[7]
>     MAD temp[9].xy, src0.xy_, src0.ww_, -src0.xy_
> 14: 
>     MAD temp[9].z, src0.__0, src0.111, src0.000
> 15: src0.xyz = temp[9]
>     DP3 temp[1].x, src0.xyz, src0.xyz
> 16: src0.xyz = temp[1]
>     MAD temp[1].x, src0.1__, src0.111, -src0.x__
> 17: src0.xyz = temp[1]
>     REPL_ALPHA temp[1].x
>     RSQ, |src0.x|
> 18: src0.xyz = temp[1]
>     REPL_ALPHA temp[9].z
>     RCP, src0.x
> 19: src0.xyz = temp[9], src1.xyz = temp[3]
>     DP3 temp[8].x, src0.xyz, src1.xyz
> 20: src0.xyz = temp[9], src1.xyz = temp[4]
>     DP3 temp[8].y, src0.xyz, src1.xyz
> 21: src0.xyz = temp[9], src1.xyz = temp[5]
>     DP3 temp[8].z, src0.xyz, src1.xyz
> 22: src0.xyz = temp[2], src1.xyz = temp[8]
>     DP3 temp[9].x, src0.xyz, src1.xyz
> 23: src0.xyz = temp[9], src1.xyz = const[7]
>     MAD temp[9].x, src0.x__, src1.x__, src0.000
> 24: src0.xyz = temp[8], src1.xyz = temp[9], src2.xyz = temp[2]
>     MAD temp[1].xyz, -src0.xyz, src1.xxx, src2.xyz
> 25: BEGIN_TEX;
> 26: TEX temp[9].xyz, temp[1].xyz_, CUBE[3];
> 27: TEX temp[8].xyz, temp[6].xy__, 2D[2];
> 28: src0.xyz = temp[8], src1.xyz = const[4]
>     MAD temp[8].xyz, src0.xyz, src1.xyz, src0.000
> 29: src0.xyz = temp[9], src1.xyz = temp[8]
>     MAD temp[1].xyz, src0.xyz, src0.111, -src1.xyz
> 30: src0.xyz = temp[1], src1.xyz = const[8], src2.xyz = temp[8]
>     MAD temp[1].xyz, src0.xyz, src1.xxx, src2.xyz
> 31: src0.xyz = const[5], src1.xyz = temp[1]
>     MAD temp[2].xyz, src0.xyz, src0.111, -src1.xyz
> 32: src0.xyz = temp[0], src1.xyz = temp[2], src2.xyz = temp[1]
>     MAD color[0].xyz, src0.xxx, src1.xyz, src2.xyz
>pc=41*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 21, tex_end: 1  (code_addr: 00420540)
>  TEX:
>    TEX t8, t1, texture[1]   (00008a01)
>    TEX t7, t1, texture[0]   (000081c1)
>  0: xyz:  t8  t7  t0 -> t1.y                 (010401c8)
>       w:  t8  t7  t0 -> t1.w                 (008401c8)
>     xyz:   t8.xyz      1.0  -t7.xyz    op: 00090a80
>       w:     t8.w      1.0    -t7.w    op: 000a8889
>  1: xyz:  c0  t1  t7 -> t9.y                 (01247060)
>       w:  c0  t1  t7 -> t9.w                 (00a47060)
>     xyz:   c0.xyz   t1.xyz   t7.xyz    op: 00020200
>       w:     c0.w     t1.w     t7.w    op: 0002c509
>  2: xyz:  c7  t0  t0 -> t9.x                 (00a40027)
>       w:  t9  t2  t0 -> t7.w                 (009c0089)
>     xyz:   t9.www      1.0      0.0    op: 00050a8c
>       w:     t2.w     c7.z      0.0    op: 4004010a
>  3: xyz:  t9  c7  t1 -> t9.xy                (01a419c9)
>       w:  t0  t0  t0 -> o0.w                 (01000000)
>     xyz:   t9.xyz   c7.xxx     -1.0    op: 000d4280
>       w:     t1.x      1.0     t1.y    op: 0001c886
>  4: xyz:  t9  c2  t0 -> t9.xy                (01a40889)
>       w:  t6  t0  t0 -> t10.w                (00a80006)
>     xyz:   t9.xyz   c2.zzz      0.0    op: 00050380
>       w:     t6.w     t9.x     t9.x    op: 05000009
>  5: xyz:  t6  t0  t0 -> t10.xy               (01a80006)
>       w: t10  t0  t0 ->                      (0000000a)
>     xyz:   t6.xyz  t10.www      0.0    op: 00050600
>       w:     t6.x     t6.x     t6.x    op: 00000000
>  6: xyz: t10  c9 c10 -> t10.xy               (01aaaa4a)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:  t10.xyz   c9.xyz  c10.xyz    op: 00020200
>       w:    t10.x    t10.x    t10.x    op: 00000000
>  7: xyz:  c6 t10  t0 -> t1.y                 (010402a6)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   c6.xyz      1.0 -t10.xyz    op: 00090a80
>       w:     c6.x     c6.x     c6.x    op: 00000000
>  8: xyz: t10  t0  t0 -> t1.x                 (0084000a)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:  t10.xyz      1.0      0.0    op: 00050a80
>       w:    t10.x    t10.x    t10.x    op: 00000000
>  9: xyz:  t1 c11  t0 -> t6.xy                (01980ac1)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz  c11.xyz      0.0    op: 00050200
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 10: xyz:  t9  t0  t0 -> t9.xy                (01a40009)
>       w:  t7  t0  t0 ->                      (00000007)
>     xyz:   t9.xyz   t7.www  -t9.xyz    op: 00080600
>       w:     t9.x     t9.x     t9.x    op: 00000000
> 11: xyz:  t0  t0  t0 -> t9.z                 (02240000)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:      0.0      1.0      0.0    op: 00050a94
>       w:     t0.x     t0.x     t0.x    op: 00000000
> 12: xyz:  t9  t0  t0 -> t1.x                 (00840009)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t9.xyz   t9.xyz   t9.xxx    op: 00804000
>       w:     t9.x     t9.x     t9.x    op: 00000000
> 13: xyz:  t1  t0  t0 -> t1.x                 (00840001)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:      1.0      1.0  -t1.xyz    op: 00080a95
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 14: xyz:  t1  t0  t0 -> t1.x                 (00840001)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xxx   t1.xxx   t1.xxx    op: 05004081
>       w:   |t1.x|     t1.x     t1.x    op: 05800040
> 15: xyz:  t1  t0  t0 -> t9.z                 (02240001)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xxx   t1.xxx   t1.xxx    op: 05004081
>       w:     t1.x     t1.x     t1.x    op: 05000000
> 16: xyz:  t9  t3  t0 -> t8.x                 (00a000c9)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t9.xyz   t3.xyz   t9.xxx    op: 00804200
>       w:     t9.x     t9.x     t9.x    op: 00000000
> 17: xyz:  t9  t4  t0 -> t8.y                 (01200109)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t9.xyz   t4.xyz   t9.xxx    op: 00804200
>       w:     t9.x     t9.x     t9.x    op: 00000000
> 18: xyz:  t9  t5  t0 -> t8.z                 (02200149)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t9.xyz   t5.xyz   t9.xxx    op: 00804200
>       w:     t9.x     t9.x     t9.x    op: 00000000
> 19: xyz:  t2  t8  t0 -> t9.x                 (00a40202)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz   t8.xyz   t2.xxx    op: 00804200
>       w:     t2.x     t2.x     t2.x    op: 00000000
> 20: xyz:  t9  c7  t0 -> t9.x                 (00a409c9)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t9.xyz   c7.xyz      0.0    op: 00050200
>       w:     t9.x     t9.x     t9.x    op: 00000000
> 21: xyz:  t8  t9  t2 -> t1.xyz               (03842248)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:  -t8.xyz   t9.xxx   t2.xyz    op: 000202a0
>       w:     t8.x     t8.x     t8.x    op: 00000000
>NODE 1: alu_offset: 22, tex_offset: 2, alu_end: 4, tex_end: 1  (code_addr: 00422116)
>  TEX:
>    TEX t9, t1, texture[3]   (00009a41)
>    TEX t8, t6, texture[2]   (00009206)
> 22: xyz:  t8  c4  t0 -> t8.xyz               (03a00908)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t8.xyz   c4.xyz      0.0    op: 00050200
>       w:     t8.x     t8.x     t8.x    op: 00000000
> 23: xyz:  t9  t8  t0 -> t1.xyz               (03840209)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t9.xyz      1.0  -t8.xyz    op: 00090a80
>       w:     t9.x     t9.x     t9.x    op: 00000000
> 24: xyz:  t1  c8  t8 -> t1.xyz               (03848a01)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   c8.xxx   t8.xyz    op: 00020280
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 25: xyz:  c5  t1  t0 -> t2.xyz               (03880065)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   c5.xyz      1.0  -t1.xyz    op: 00090a80
>       w:     c5.x     c5.x     c5.x    op: 00000000
> 26: xyz:  t0  t2  t1 -> o0.xyz               (1c001080)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xxx   t2.xyz   t1.xyz    op: 00020201
>       w:     t0.x     t0.x     t0.x    op: 00000000
>r300: Initial fragment program
>FRAG
>DCL IN[0], GENERIC[0], PERSPECTIVE
>DCL IN[1], GENERIC[2], PERSPECTIVE
>DCL IN[2], GENERIC[3], PERSPECTIVE
>DCL IN[3], GENERIC[4], PERSPECTIVE
>DCL IN[4], GENERIC[5], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL SAMP[1]
>DCL SAMP[2]
>DCL CONST[1]
>DCL TEMP[0..4]
>IMM FLT32 {    2.0000,     1.0000,     0.0000,     0.0000}
>  0: TEX TEMP[0], IN[0], SAMP[0], 2D
>  1: MOV TEMP[0].x, TEMP[0].wwww
>  2: MAD TEMP[1], TEMP[0], IMM[0].xxxx, -IMM[0].yyyy
>  3: MOV TEMP[1].z, IMM[0].zzzz
>  4: DP3 TEMP[0].x, TEMP[1], TEMP[1]
>  5: ADD TEMP[0].x, IMM[0].yyyy, -TEMP[0].xxxx
>  6: RSQ TEMP[0].x, TEMP[0].xxxx
>  7: RCP TEMP[1].z, TEMP[0].xxxx
>  8: DP3 TEMP[0].x, TEMP[1], IN[2]
>  9: DP3 TEMP[0].y, TEMP[1], IN[3]
> 10: DP3 TEMP[0].z, TEMP[1], IN[4]
> 11: DP3 TEMP[2].w, IN[1], IN[1]
> 12: RSQ TEMP[2].w, TEMP[2].wwww
> 13: MUL TEMP[2], IN[1], TEMP[2].wwww
> 14: DP3 TEMP[1], TEMP[2], TEMP[0]
> 15: MUL TEMP[1], TEMP[1], IMM[0].xxxx
> 16: MAD TEMP[3], -TEMP[0], TEMP[1].xxxx, TEMP[2]
> 17: DP3_SAT TEMP[0], -TEMP[2], TEMP[0]
> 18: SUB TEMP[0], IMM[0].yyyy, TEMP[0]
> 19: POW TEMP[4], TEMP[0].xxxx, CONST[1].xxxx
> 20: TEX TEMP[3], TEMP[3], SAMP[1], CUBE
> 21: TEX TEMP[1], IN[0], SAMP[2], 2D
> 22: MUL OUT[0].xyz, TEMP[3], CONST[1].yyyy
> 23: ADD TEMP[4], TEMP[4], CONST[1].zzzz
> 24: MUL OUT[0].w, TEMP[1].xxxx, TEMP[4]
> 25: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: TEX temp[0], input[0], 2D[0];
>  1: MOV temp[0].x, temp[0].wwww;
>  2: MAD temp[1], temp[0], const[2].xxxx, -const[2].yyyy;
>  3: MOV temp[1].z, const[2].zzzz;
>  4: DP3 temp[0].x, temp[1], temp[1];
>  5: ADD temp[0].x, const[2].yyyy, -temp[0].xxxx;
>  6: RSQ temp[0].x, temp[0].xxxx;
>  7: RCP temp[1].z, temp[0].xxxx;
>  8: DP3 temp[0].x, temp[1], input[2];
>  9: DP3 temp[0].y, temp[1], input[3];
> 10: DP3 temp[0].z, temp[1], input[4];
> 11: DP3 temp[2].w, input[1], input[1];
> 12: RSQ temp[2].w, temp[2].wwww;
> 13: MUL temp[2], input[1], temp[2].wwww;
> 14: DP3 temp[1], temp[2], temp[0];
> 15: MUL temp[1], temp[1], const[2].xxxx;
> 16: MAD temp[3], -temp[0], temp[1].xxxx, temp[2];
> 17: DP3_SAT temp[0], -temp[2], temp[0];
> 18: SUB temp[0], const[2].yyyy, temp[0];
> 19: POW temp[4], temp[0].xxxx, const[1].xxxx;
> 20: TEX temp[3], temp[3], CUBE[1];
> 21: TEX temp[1], input[0], 2D[2];
> 22: MUL output[0].xyz, temp[3], const[1].yyyy;
> 23: ADD temp[4], temp[4], const[1].zzzz;
> 24: MUL output[0].w, temp[1].xxxx, temp[4];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: TEX temp[0], input[0], 2D[0];
>  1: MOV temp[0].x, temp[0].wwww;
>  2: MAD temp[1], temp[0], const[2].xxxx, -const[2].yyyy;
>  3: MOV temp[1].z, const[2].zzzz;
>  4: DP3 temp[0].x, temp[1], temp[1];
>  5: ADD temp[0].x, const[2].yyyy, -temp[0].xxxx;
>  6: RSQ temp[0].x, temp[0].xxxx;
>  7: RCP temp[1].z, temp[0].xxxx;
>  8: DP3 temp[0].x, temp[1], input[2];
>  9: DP3 temp[0].y, temp[1], input[3];
> 10: DP3 temp[0].z, temp[1], input[4];
> 11: DP3 temp[2].w, input[1], input[1];
> 12: RSQ temp[2].w, temp[2].wwww;
> 13: MUL temp[2], input[1], temp[2].wwww;
> 14: DP3 temp[1], temp[2], temp[0];
> 15: MUL temp[1], temp[1], const[2].xxxx;
> 16: MAD temp[3], -temp[0], temp[1].xxxx, temp[2];
> 17: DP3_SAT temp[0], -temp[2], temp[0];
> 18: SUB temp[0], const[2].yyyy, temp[0];
> 19: POW temp[4], temp[0].xxxx, const[1].xxxx;
> 20: TEX temp[3], temp[3], CUBE[1];
> 21: TEX temp[1], input[0], 2D[2];
> 22: MUL output[0].xyz, temp[3], const[1].yyyy;
> 23: ADD temp[4], temp[4], const[1].zzzz;
> 24: MUL output[0].w, temp[1].xxxx, temp[4];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: TEX temp[0], input[0], 2D[0];
>  1: MOV temp[0].x, temp[0].wwww;
>  2: MAD temp[1], temp[0], const[2].xxxx, -const[2].yyyy;
>  3: MOV temp[1].z, const[2].zzzz;
>  4: DP3 temp[0].x, temp[1], temp[1];
>  5: ADD temp[0].x, const[2].yyyy, -temp[0].xxxx;
>  6: RSQ temp[0].x, temp[0].xxxx;
>  7: RCP temp[1].z, temp[0].xxxx;
>  8: DP3 temp[0].x, temp[1], input[2];
>  9: DP3 temp[0].y, temp[1], input[3];
> 10: DP3 temp[0].z, temp[1], input[4];
> 11: DP3 temp[2].w, input[1], input[1];
> 12: RSQ temp[2].w, temp[2].wwww;
> 13: MUL temp[2], input[1], temp[2].wwww;
> 14: DP3 temp[1], temp[2], temp[0];
> 15: MUL temp[1], temp[1], const[2].xxxx;
> 16: MAD temp[3], -temp[0], temp[1].xxxx, temp[2];
> 17: DP3_SAT temp[0], -temp[2], temp[0];
> 18: SUB temp[0], const[2].yyyy, temp[0];
> 19: POW temp[4], temp[0].xxxx, const[1].xxxx;
> 20: TEX temp[3], temp[3], CUBE[1];
> 21: TEX temp[1], input[0], 2D[2];
> 22: MUL output[0].xyz, temp[3], const[1].yyyy;
> 23: ADD temp[4], temp[4], const[1].zzzz;
> 24: MUL output[0].w, temp[1].xxxx, temp[4];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: TEX temp[0], input[0], 2D[0];
>  1: MOV temp[0].x, temp[0].wwww;
>  2: MAD temp[1], temp[0], const[2].xxxx, -const[2].yyyy;
>  3: MOV temp[1].z, const[2].zzzz;
>  4: DP3 temp[0].x, temp[1], temp[1];
>  5: ADD temp[0].x, const[2].yyyy, -temp[0].xxxx;
>  6: RSQ temp[0].x, |temp[0].xxxx|;
>  7: RCP temp[1].z, temp[0].xxxx;
>  8: DP3 temp[0].x, temp[1], input[2];
>  9: DP3 temp[0].y, temp[1], input[3];
> 10: DP3 temp[0].z, temp[1], input[4];
> 11: DP3 temp[2].w, input[1], input[1];
> 12: RSQ temp[2].w, |temp[2].wwww|;
> 13: MUL temp[2], input[1], temp[2].wwww;
> 14: DP3 temp[1], temp[2], temp[0];
> 15: MUL temp[1], temp[1], const[2].xxxx;
> 16: MAD temp[3], -temp[0], temp[1].xxxx, temp[2];
> 17: DP3_SAT temp[0], -temp[2], temp[0];
> 18: ADD temp[0], const[2].yyyy, -temp[0];
> 19: LG2 temp[5].w, temp[0].xxxx;
> 20: MUL temp[5].w, temp[5].wwww, const[1].xxxx;
> 21: EX2 temp[4], temp[5].wwww;
> 22: TEX temp[3], temp[3], CUBE[1];
> 23: TEX temp[1], input[0], 2D[2];
> 24: MUL output[0].xyz, temp[3], const[1].yyyy;
> 25: ADD temp[4], temp[4], const[1].zzzz;
> 26: MUL output[0].w, temp[1].xxxx, temp[4];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: TEX temp[0], input[0], 2D[0];
>  1: MOV temp[0].x, temp[0].wwww;
>  2: MAD temp[1], temp[0], const[2].xxxx, -const[2].yyyy;
>  3: MOV temp[1].z, const[2].zzzz;
>  4: DP3 temp[0].x, temp[1], temp[1];
>  5: ADD temp[0].x, const[2].yyyy, -temp[0].xxxx;
>  6: RSQ temp[0].x, |temp[0].xxxx|;
>  7: RCP temp[1].z, temp[0].xxxx;
>  8: DP3 temp[0].x, temp[1], input[2];
>  9: DP3 temp[0].y, temp[1], input[3];
> 10: DP3 temp[0].z, temp[1], input[4];
> 11: DP3 temp[2].w, input[1], input[1];
> 12: RSQ temp[2].w, |temp[2].wwww|;
> 13: MUL temp[2], input[1], temp[2].wwww;
> 14: DP3 temp[1], temp[2], temp[0];
> 15: MUL temp[1], temp[1], const[2].xxxx;
> 16: MAD temp[3], -temp[0], temp[1].xxxx, temp[2];
> 17: DP3_SAT temp[0], -temp[2], temp[0];
> 18: ADD temp[0], const[2].yyyy, -temp[0];
> 19: LG2 temp[5].w, temp[0].xxxx;
> 20: MUL temp[5].w, temp[5].wwww, const[1].xxxx;
> 21: EX2 temp[4], temp[5].wwww;
> 22: TEX temp[3], temp[3], CUBE[1];
> 23: MUL temp[6], input[0], const[3];
> 24: TEX temp[1], temp[6], 2D[2];
> 25: MUL output[0].xyz, temp[3], const[1].yyyy;
> 26: ADD temp[4], temp[4], const[1].zzzz;
> 27: MUL output[0].w, temp[1].xxxx, temp[4];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: TEX temp[0].yw, input[0].xy__, 2D[0];
>  1: MOV temp[0].x, temp[0].w___;
>  2: MAD temp[1].xy, temp[0].xy__, const[2].xx__, -const[2].yy__;
>  3: MOV temp[1].z, const[2].__z_;
>  4: DP3 temp[0].x, temp[1].xyz_, temp[1].xyz_;
>  5: ADD temp[0].x, const[2].y___, -temp[0].x___;
>  6: RSQ temp[0].x, |temp[0].x___|;
>  7: RCP temp[1].z, temp[0].x___;
>  8: DP3 temp[0].x, temp[1].xyz_, input[2].xyz_;
>  9: DP3 temp[0].y, temp[1].xyz_, input[3].xyz_;
> 10: DP3 temp[0].z, temp[1].xyz_, input[4].xyz_;
> 11: DP3 temp[2].w, input[1].xyz_, input[1].xyz_;
> 12: RSQ temp[2].w, |temp[2].w___|;
> 13: MUL temp[2].xyz, input[1].xyz_, temp[2].www_;
> 14: DP3 temp[1].x, temp[2].xyz_, temp[0].xyz_;
> 15: MUL temp[1].x, temp[1].x___, const[2].x___;
> 16: MAD temp[3].xyz, -temp[0].xyz_, temp[1].xxx_, temp[2].xyz_;
> 17: DP3_SAT temp[0].x, -temp[2].xyz_, temp[0].xyz_;
> 18: ADD temp[0].x, const[2].y___, -temp[0].x___;
> 19: LG2 temp[5].w, temp[0].x___;
> 20: MUL temp[5].w, temp[5].___w, const[1].___x;
> 21: EX2 temp[4].w, temp[5].w___;
> 22: TEX temp[3].xyz, temp[3].xyz_, CUBE[1];
> 23: MUL temp[6].xy, input[0].xy__, const[3].xy__;
> 24: TEX temp[1].x, temp[6].xy__, 2D[2];
> 25: MUL output[0].xyz, temp[3].xyz_, const[1].yyy_;
> 26: ADD temp[4].w, temp[4].___w, const[1].___z;
> 27: MUL output[0].w, temp[1].___x, temp[4].___w;
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: TEX temp[0].yw, input[0].xy__, 2D[0];
>  1: MOV temp[0].x, temp[0].w___;
>  2: MAD temp[1].xy, temp[0].xy__, const[2].xx__, -none.11__;
>  3: MOV temp[1].z, none.__0_;
>  4: DP3 temp[0].x, temp[1].xyz_, temp[1].xyz_;
>  5: ADD temp[0].x, none.1___, -temp[0].x___;
>  6: RSQ temp[0].x, |temp[0].x___|;
>  7: RCP temp[1].z, temp[0].x___;
>  8: DP3 temp[0].x, temp[1].xyz_, input[2].xyz_;
>  9: DP3 temp[0].y, temp[1].xyz_, input[3].xyz_;
> 10: DP3 temp[0].z, temp[1].xyz_, input[4].xyz_;
> 11: DP3 temp[2].w, input[1].xyz_, input[1].xyz_;
> 12: RSQ temp[2].w, |temp[2].w___|;
> 13: MUL temp[2].xyz, input[1].xyz_, temp[2].www_;
> 14: DP3 temp[1].x, temp[2].xyz_, temp[0].xyz_;
> 15: MUL temp[1].x, temp[1].x___, const[2].x___;
> 16: MAD temp[3].xyz, -temp[0].xyz_, temp[1].xxx_, temp[2].xyz_;
> 17: DP3_SAT temp[0].x, -temp[2].xyz_, temp[0].xyz_;
> 18: ADD temp[0].x, none.1___, -temp[0].x___;
> 19: LG2 temp[5].w, temp[0].x___;
> 20: MUL temp[5].w, temp[5].___w, const[1].___x;
> 21: EX2 temp[4].w, temp[5].w___;
> 22: TEX temp[3].xyz, temp[3].xyz_, CUBE[1];
> 23: MUL temp[6].xy, input[0].xy__, const[3].xy__;
> 24: TEX temp[1].x, temp[6].xy__, 2D[2];
> 25: MUL output[0].xyz, temp[3].xyz_, const[1].yyy_;
> 26: ADD temp[4].w, temp[4].___w, const[1].___z;
> 27: MUL output[0].w, temp[1].___x, temp[4].___w;
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: TEX temp[0].yw, input[0].xy__, 2D[0];
>  1: MOV temp[0].x, temp[0].w___;
>  2: MAD temp[1].xy, temp[0].xy__, const[2].xx__, -none.11__;
>  3: MOV temp[1].z, none.__0_;
>  4: DP3 temp[0].x, temp[1].xyz_, temp[1].xyz_;
>  5: ADD temp[0].x, none.1___, -temp[0].x___;
>  6: RSQ temp[0].x, |temp[0].x___|;
>  7: RCP temp[1].z, temp[0].x___;
>  8: DP3 temp[0].x, temp[1].xyz_, input[2].xyz_;
>  9: DP3 temp[0].y, temp[1].xyz_, input[3].xyz_;
> 10: DP3 temp[0].z, temp[1].xyz_, input[4].xyz_;
> 11: DP3 temp[2].w, input[1].xyz_, input[1].xyz_;
> 12: RSQ temp[2].w, |temp[2].w___|;
> 13: MUL temp[2].xyz, input[1].xyz_, temp[2].www_;
> 14: DP3 temp[1].x, temp[2].xyz_, temp[0].xyz_;
> 15: MUL temp[1].x, temp[1].x___, const[2].x___;
> 16: MAD temp[3].xyz, -temp[0].xyz_, temp[1].xxx_, temp[2].xyz_;
> 17: DP3_SAT temp[0].x, -temp[2].xyz_, temp[0].xyz_;
> 18: ADD temp[0].x, none.1___, -temp[0].x___;
> 19: LG2 temp[5].w, temp[0].x___;
> 20: MUL temp[5].w, temp[5].___w, const[1].___x;
> 21: EX2 temp[4].w, temp[5].w___;
> 22: TEX temp[3].xyz, temp[3].xyz_, CUBE[1];
> 23: MUL temp[6].xy, input[0].xy__, const[3].xy__;
> 24: TEX temp[1].x, temp[6].xy__, 2D[2];
> 25: MUL output[0].xyz, temp[3].xyz_, const[1].yyy_;
> 26: ADD temp[4].w, temp[4].___w, const[1].___z;
> 27: MUL output[0].w, temp[1].___x, temp[4].___w;
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: TEX temp[0].yw, input[0].xy__, 2D[0];
>  1: src0.w = temp[0]
>     MAD temp[0].x, src0.w__, src0.111, src0.000
>  2: src0.xyz = temp[0], src1.xyz = const[2]
>     MAD temp[1].xy, src0.xy_, src1.xx_, -src0.11_
>  3: 
>     MAD temp[1].z, src0.__0, src0.111, src0.000
>  4: src0.xyz = temp[1]
>     DP3 temp[0].x, src0.xyz, src0.xyz
>  5: src0.xyz = temp[0]
>     MAD temp[0].x, src0.1__, src0.111, -src0.x__
>  6: src0.xyz = temp[0]
>     REPL_ALPHA temp[0].x
>     RSQ, |src0.x|
>  7: src0.xyz = temp[0]
>     REPL_ALPHA temp[1].z
>     RCP, src0.x
>  8: src0.xyz = temp[1], src1.xyz = input[2]
>     DP3 temp[0].x, src0.xyz, src1.xyz
>  9: src0.xyz = temp[1], src1.xyz = input[3]
>     DP3 temp[0].y, src0.xyz, src1.xyz
> 10: src0.xyz = temp[1], src1.xyz = input[4]
>     DP3 temp[0].z, src0.xyz, src1.xyz
> 11: src0.xyz = input[1]
>     DP3, src0.xyz, src0.xyz
>     DP3 temp[2].w, src0._, src0._
> 12: src0.w = temp[2]
>     RSQ temp[2].w, |src0.w|
> 13: src0.xyz = input[1], src0.w = temp[2]
>     MAD temp[2].xyz, src0.xyz, src0.www, src0.000
> 14: src0.xyz = temp[2], src1.xyz = temp[0]
>     DP3 temp[1].x, src0.xyz, src1.xyz
> 15: src0.xyz = temp[1], src1.xyz = const[2]
>     MAD temp[1].x, src0.x__, src1.x__, src0.000
> 16: src0.xyz = temp[0], src1.xyz = temp[1], src2.xyz = temp[2]
>     MAD temp[3].xyz, -src0.xyz, src1.xxx, src2.xyz
> 17: src0.xyz = temp[2], src1.xyz = temp[0]
>     DP3_SAT temp[0].x, -src0.xyz, src1.xyz
> 18: src0.xyz = temp[0]
>     MAD temp[0].x, src0.1__, src0.111, -src0.x__
> 19: src0.xyz = temp[0]
>     LG2 temp[5].w, src0.x
> 20: src0.xyz = const[1], src0.w = temp[5]
>     MAD temp[5].w, src0.w, src0.x, src0.0
> 21: src0.w = temp[5]
>     EX2 temp[4].w, src0.w
> 22: TEX temp[3].xyz, temp[3].xyz_, CUBE[1];
> 23: src0.xyz = input[0], src1.xyz = const[3]
>     MAD temp[6].xy, src0.xy_, src1.xy_, src0.000
> 24: TEX temp[1].x, temp[6].xy__, 2D[2];
> 25: src0.xyz = temp[3], src1.xyz = const[1]
>     MAD color[0].xyz, src0.xyz, src1.yyy, src0.000
> 26: src0.xyz = const[1], src0.w = temp[4]
>     MAD temp[4].w, src0.w, src0.1, src0.z
> 27: src0.xyz = temp[1], src0.w = temp[4]
>     MAD color[0].w, src0.x, src0.w, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[0].yw, input[0].xy__, 2D[0];
>  2: src0.xyz = input[1]
>     DP3, src0.xyz, src0.xyz
>     DP3 temp[2].w, src0._, src0._
>  3: src0.w = temp[0], src1.w = temp[2]
>     MAD temp[0].x, src0.w__, src0.111, src0.000
>     RSQ temp[2].w, |src1.w|
>  4: src0.xyz = input[1], src0.w = temp[2]
>     MAD temp[2].xyz, src0.xyz, src0.www, src0.000
>  5: src0.xyz = temp[0], src1.xyz = const[2]
>     MAD temp[1].xy, src0.xy_, src1.xx_, -src0.11_
>  6: src0.xyz = input[0], src1.xyz = const[3]
>     MAD temp[6].xy, src0.xy_, src1.xy_, src0.000
>  7: 
>     MAD temp[1].z, src0.__0, src0.111, src0.000
>  8: src0.xyz = temp[1]
>     DP3 temp[0].x, src0.xyz, src0.xyz
>  9: src0.xyz = temp[0]
>     MAD temp[0].x, src0.1__, src0.111, -src0.x__
> 10: src0.xyz = temp[0]
>     REPL_ALPHA temp[0].x
>     RSQ, |src0.x|
> 11: src0.xyz = temp[0]
>     REPL_ALPHA temp[1].z
>     RCP, src0.x
> 12: src0.xyz = temp[1], src1.xyz = input[2]
>     DP3 temp[0].x, src0.xyz, src1.xyz
> 13: src0.xyz = temp[1], src1.xyz = input[3]
>     DP3 temp[0].y, src0.xyz, src1.xyz
> 14: src0.xyz = temp[1], src1.xyz = input[4]
>     DP3 temp[0].z, src0.xyz, src1.xyz
> 15: src0.xyz = temp[2], src1.xyz = temp[0]
>     DP3 temp[1].x, src0.xyz, src1.xyz
> 16: src0.xyz = temp[1], src1.xyz = const[2]
>     MAD temp[1].x, src0.x__, src1.x__, src0.000
> 17: src0.xyz = temp[0], src1.xyz = temp[1], src2.xyz = temp[2]
>     MAD temp[3].xyz, -src0.xyz, src1.xxx, src2.xyz
> 18: src0.xyz = temp[2], src1.xyz = temp[0]
>     DP3_SAT temp[0].x, -src0.xyz, src1.xyz
> 19: src0.xyz = temp[0]
>     MAD temp[0].x, src0.1__, src0.111, -src0.x__
> 20: src0.xyz = temp[0]
>     LG2 temp[5].w, src0.x
> 21: src0.xyz = const[1], src0.w = temp[5]
>     MAD temp[5].w, src0.w, src0.x, src0.0
> 22: src0.w = temp[5]
>     EX2 temp[4].w, src0.w
> 23: src0.xyz = const[1], src0.w = temp[4]
>     MAD temp[4].w, src0.w, src0.1, src0.z
> 24: BEGIN_TEX;
> 25: TEX temp[3].xyz, temp[3].xyz_, CUBE[1];
> 26: TEX temp[1].x, temp[6].xy__, 2D[2];
> 27: src0.xyz = temp[3], src0.w = temp[4], src1.xyz = const[1], src2.xyz = temp[1]
>     MAD color[0].xyz, src0.xyz, src1.yyy, src0.000
>     MAD color[0].w, src2.x, src0.w, src0.0
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[5].yw, temp[0].xy__, 2D[0];
>  2: src0.xyz = temp[1]
>     DP3, src0.xyz, src0.xyz
>     DP3 temp[6].w, src0._, src0._
>  3: src0.w = temp[5], src1.w = temp[6]
>     MAD temp[5].x, src0.w__, src0.111, src0.000
>     RSQ temp[6].w, |src1.w|
>  4: src0.xyz = temp[1], src0.w = temp[6]
>     MAD temp[6].xyz, src0.xyz, src0.www, src0.000
>  5: src0.xyz = temp[5], src1.xyz = const[2]
>     MAD temp[1].xy, src0.xy_, src1.xx_, -src0.11_
>  6: src0.xyz = temp[0], src1.xyz = const[3]
>     MAD temp[7].xy, src0.xy_, src1.xy_, src0.000
>  7: 
>     MAD temp[1].z, src0.__0, src0.111, src0.000
>  8: src0.xyz = temp[1]
>     DP3 temp[5].x, src0.xyz, src0.xyz
>  9: src0.xyz = temp[5]
>     MAD temp[5].x, src0.1__, src0.111, -src0.x__
> 10: src0.xyz = temp[5]
>     REPL_ALPHA temp[5].x
>     RSQ, |src0.x|
> 11: src0.xyz = temp[5]
>     REPL_ALPHA temp[1].z
>     RCP, src0.x
> 12: src0.xyz = temp[1], src1.xyz = temp[2]
>     DP3 temp[5].x, src0.xyz, src1.xyz
> 13: src0.xyz = temp[1], src1.xyz = temp[3]
>     DP3 temp[5].y, src0.xyz, src1.xyz
> 14: src0.xyz = temp[1], src1.xyz = temp[4]
>     DP3 temp[5].z, src0.xyz, src1.xyz
> 15: src0.xyz = temp[6], src1.xyz = temp[5]
>     DP3 temp[1].x, src0.xyz, src1.xyz
> 16: src0.xyz = temp[1], src1.xyz = const[2]
>     MAD temp[1].x, src0.x__, src1.x__, src0.000
> 17: src0.xyz = temp[5], src1.xyz = temp[1], src2.xyz = temp[6]
>     MAD temp[0].xyz, -src0.xyz, src1.xxx, src2.xyz
> 18: src0.xyz = temp[6], src1.xyz = temp[5]
>     DP3_SAT temp[5].x, -src0.xyz, src1.xyz
> 19: src0.xyz = temp[5]
>     MAD temp[5].x, src0.1__, src0.111, -src0.x__
> 20: src0.xyz = temp[5]
>     LG2 temp[2].w, src0.x
> 21: src0.xyz = const[1], src0.w = temp[2]
>     MAD temp[2].w, src0.w, src0.x, src0.0
> 22: src0.w = temp[2]
>     EX2 temp[2].w, src0.w
> 23: src0.xyz = const[1], src0.w = temp[2]
>     MAD temp[2].w, src0.w, src0.1, src0.z
> 24: BEGIN_TEX;
> 25: TEX temp[0].xyz, temp[0].xyz_, CUBE[1];
> 26: TEX temp[1].x, temp[7].xy__, 2D[2];
> 27: src0.xyz = temp[0], src0.w = temp[2], src1.xyz = const[1], src2.xyz = temp[1]
>     MAD color[0].xyz, src0.xyz, src1.yyy, src0.000
>     MAD color[0].w, src2.x, src0.w, src0.0
>pc=42*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 21, tex_end: 0  (code_addr: 00000540)
>  TEX:
>    TEX t5, t0, texture[0]   (00008140)
>  0: xyz:  t1  t0  t0 ->                      (00000001)
>       w:  t0  t0  t0 -> t6.w                 (00980000)
>     xyz:   t1.xyz   t1.xyz   t1.xxx    op: 00804000
>       w:      1.0      1.0     t1.x    op: 00800891
>  1: xyz:  t0  t0  t0 -> t5.x                 (00940000)
>       w:  t5  t6  t0 -> t6.w                 (00980185)
>     xyz:   t5.www      1.0      0.0    op: 00050a8c
>       w:   |t6.w|     t0.x     t0.x    op: 0580004a
>  2: xyz:  t1  t0  t0 -> t6.xyz               (03980001)
>       w:  t6  t0  t0 ->                      (00000006)
>     xyz:   t1.xyz   t6.www      0.0    op: 00050600
>       w:     t1.x     t1.x     t1.x    op: 00000000
>  3: xyz:  t5  c2  t0 -> t1.xy                (01840885)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t5.xyz   c2.xxx     -1.0    op: 000d4280
>       w:     t5.x     t5.x     t5.x    op: 00000000
>  4: xyz:  t0  c3  t0 -> t7.xy                (019c08c0)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xyz   c3.xyz      0.0    op: 00050200
>       w:     t0.x     t0.x     t0.x    op: 00000000
>  5: xyz:  t0  t0  t0 -> t1.z                 (02040000)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:      0.0      1.0      0.0    op: 00050a94
>       w:     t0.x     t0.x     t0.x    op: 00000000
>  6: xyz:  t1  t0  t0 -> t5.x                 (00940001)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   t1.xyz   t1.xxx    op: 00804000
>       w:     t1.x     t1.x     t1.x    op: 00000000
>  7: xyz:  t5  t0  t0 -> t5.x                 (00940005)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:      1.0      1.0  -t5.xyz    op: 00080a95
>       w:     t5.x     t5.x     t5.x    op: 00000000
>  8: xyz:  t5  t0  t0 -> t5.x                 (00940005)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t5.xxx   t5.xxx   t5.xxx    op: 05004081
>       w:   |t5.x|     t5.x     t5.x    op: 05800040
>  9: xyz:  t5  t0  t0 -> t1.z                 (02040005)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t5.xxx   t5.xxx   t5.xxx    op: 05004081
>       w:     t5.x     t5.x     t5.x    op: 05000000
> 10: xyz:  t1  t2  t0 -> t5.x                 (00940081)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   t2.xyz   t1.xxx    op: 00804200
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 11: xyz:  t1  t3  t0 -> t5.y                 (011400c1)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   t3.xyz   t1.xxx    op: 00804200
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 12: xyz:  t1  t4  t0 -> t5.z                 (02140101)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   t4.xyz   t1.xxx    op: 00804200
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 13: xyz:  t6  t5  t0 -> t1.x                 (00840146)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t6.xyz   t5.xyz   t6.xxx    op: 00804200
>       w:     t6.x     t6.x     t6.x    op: 00000000
> 14: xyz:  t1  c2  t0 -> t1.x                 (00840881)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   c2.xyz      0.0    op: 00050200
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 15: xyz:  t5  t1  t6 -> t0.xyz               (03806045)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:  -t5.xyz   t1.xxx   t6.xyz    op: 000202a0
>       w:     t5.x     t5.x     t5.x    op: 00000000
> 16: xyz:  t6  t5  t0 -> t5.x                 (00940146)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:  -t6.xyz   t5.xyz   t6.xxx    op: 40804220
>       w:     t6.x     t6.x     t6.x    op: 00000000
> 17: xyz:  t5  t0  t0 -> t5.x                 (00940005)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:      1.0      1.0  -t5.xyz    op: 00080a95
>       w:     t5.x     t5.x     t5.x    op: 00000000
> 18: xyz:  t5  t0  t0 ->                      (00000005)
>       w:  t0  t0  t0 -> t2.w                 (00880000)
>     xyz:   t5.xxx   t5.xxx   t5.xxx    op: 00004081
>       w:     t5.x     t5.x     t5.x    op: 04800000
> 19: xyz:  c1  t0  t0 ->                      (00000021)
>       w:  t2  t0  t0 -> t2.w                 (00880002)
>     xyz:   c1.xxx   c1.xxx   c1.xxx    op: 00004081
>       w:     t2.w     c1.x      0.0    op: 00040009
> 20: xyz:  t0  t0  t0 ->                      (00000000)
>       w:  t2  t0  t0 -> t2.w                 (00880002)
>     xyz:   t0.xxx   t0.xxx   t0.xxx    op: 00004081
>       w:     t2.w     t0.x     t0.x    op: 04000009
> 21: xyz:  c1  t0  t0 ->                      (00000021)
>       w:  t2  t0  t0 -> t2.w                 (00880002)
>     xyz:   c1.xxx   c1.xxx   c1.xxx    op: 00004081
>       w:     t2.w      1.0     c1.z    op: 00008889
>NODE 1: alu_offset: 22, tex_offset: 1, alu_end: 0, tex_end: 1  (code_addr: 00421016)
>  TEX:
>    TEX t0, t0, texture[1]   (00008800)
>    TEX t1, t7, texture[2]   (00009047)
> 22: xyz:  t0  c1  t1 -> o0.xyz               (1c001840)
>       w:  t2  t0  t0 -> o0.w                 (01000002)
>     xyz:   t0.xyz   c1.yyy      0.0    op: 00050300
>       w:     t1.x     t2.w      0.0    op: 00040486
>r300: Initial fragment program
>FRAG
>DCL IN[0], GENERIC[0], PERSPECTIVE
>DCL IN[1], GENERIC[2], PERSPECTIVE
>DCL IN[2], GENERIC[3], PERSPECTIVE
>DCL IN[3], GENERIC[4], PERSPECTIVE
>DCL IN[4], GENERIC[5], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL SAMP[1]
>DCL SAMP[2]
>DCL CONST[1]
>DCL TEMP[0..4]
>IMM FLT32 {    2.0000,     1.0000,     0.0000,     0.0000}
>  0: TEX TEMP[0], IN[0], SAMP[0], 2D
>  1: MOV TEMP[0].x, TEMP[0].wwww
>  2: MAD TEMP[1], TEMP[0], IMM[0].xxxx, -IMM[0].yyyy
>  3: MOV TEMP[1].z, IMM[0].zzzz
>  4: DP3 TEMP[0].x, TEMP[1], TEMP[1]
>  5: ADD TEMP[0].x, IMM[0].yyyy, -TEMP[0].xxxx
>  6: RSQ TEMP[0].x, TEMP[0].xxxx
>  7: RCP TEMP[1].z, TEMP[0].xxxx
>  8: DP3 TEMP[0].x, TEMP[1], IN[2]
>  9: DP3 TEMP[0].y, TEMP[1], IN[3]
> 10: DP3 TEMP[0].z, TEMP[1], IN[4]
> 11: DP3 TEMP[2].w, IN[1], IN[1]
> 12: RSQ TEMP[2].w, TEMP[2].wwww
> 13: MUL TEMP[2], IN[1], TEMP[2].wwww
> 14: DP3 TEMP[1], TEMP[2], TEMP[0]
> 15: MUL TEMP[1], TEMP[1], IMM[0].xxxx
> 16: MAD TEMP[3], -TEMP[0], TEMP[1].xxxx, TEMP[2]
> 17: DP3_SAT TEMP[0], -TEMP[2], TEMP[0]
> 18: SUB TEMP[0], IMM[0].yyyy, TEMP[0]
> 19: POW TEMP[4], TEMP[0].xxxx, CONST[1].xxxx
> 20: TEX TEMP[3], TEMP[3], SAMP[1], CUBE
> 21: TEX TEMP[1], IN[0], SAMP[2], 2D
> 22: MUL OUT[0].xyz, TEMP[3], CONST[1].yyyy
> 23: ADD TEMP[4], TEMP[4], CONST[1].zzzz
> 24: MUL OUT[0].w, TEMP[1].xxxx, TEMP[4]
> 25: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: TEX temp[0], input[0], 2D[0];
>  1: MOV temp[0].x, temp[0].wwww;
>  2: MAD temp[1], temp[0], const[2].xxxx, -const[2].yyyy;
>  3: MOV temp[1].z, const[2].zzzz;
>  4: DP3 temp[0].x, temp[1], temp[1];
>  5: ADD temp[0].x, const[2].yyyy, -temp[0].xxxx;
>  6: RSQ temp[0].x, temp[0].xxxx;
>  7: RCP temp[1].z, temp[0].xxxx;
>  8: DP3 temp[0].x, temp[1], input[2];
>  9: DP3 temp[0].y, temp[1], input[3];
> 10: DP3 temp[0].z, temp[1], input[4];
> 11: DP3 temp[2].w, input[1], input[1];
> 12: RSQ temp[2].w, temp[2].wwww;
> 13: MUL temp[2], input[1], temp[2].wwww;
> 14: DP3 temp[1], temp[2], temp[0];
> 15: MUL temp[1], temp[1], const[2].xxxx;
> 16: MAD temp[3], -temp[0], temp[1].xxxx, temp[2];
> 17: DP3_SAT temp[0], -temp[2], temp[0];
> 18: SUB temp[0], const[2].yyyy, temp[0];
> 19: POW temp[4], temp[0].xxxx, const[1].xxxx;
> 20: TEX temp[3], temp[3], CUBE[1];
> 21: TEX temp[1], input[0], 2D[2];
> 22: MUL output[0].xyz, temp[3], const[1].yyyy;
> 23: ADD temp[4], temp[4], const[1].zzzz;
> 24: MUL output[0].w, temp[1].xxxx, temp[4];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: TEX temp[0], input[0], 2D[0];
>  1: MOV temp[0].x, temp[0].wwww;
>  2: MAD temp[1], temp[0], const[2].xxxx, -const[2].yyyy;
>  3: MOV temp[1].z, const[2].zzzz;
>  4: DP3 temp[0].x, temp[1], temp[1];
>  5: ADD temp[0].x, const[2].yyyy, -temp[0].xxxx;
>  6: RSQ temp[0].x, temp[0].xxxx;
>  7: RCP temp[1].z, temp[0].xxxx;
>  8: DP3 temp[0].x, temp[1], input[2];
>  9: DP3 temp[0].y, temp[1], input[3];
> 10: DP3 temp[0].z, temp[1], input[4];
> 11: DP3 temp[2].w, input[1], input[1];
> 12: RSQ temp[2].w, temp[2].wwww;
> 13: MUL temp[2], input[1], temp[2].wwww;
> 14: DP3 temp[1], temp[2], temp[0];
> 15: MUL temp[1], temp[1], const[2].xxxx;
> 16: MAD temp[3], -temp[0], temp[1].xxxx, temp[2];
> 17: DP3_SAT temp[0], -temp[2], temp[0];
> 18: SUB temp[0], const[2].yyyy, temp[0];
> 19: POW temp[4], temp[0].xxxx, const[1].xxxx;
> 20: TEX temp[3], temp[3], CUBE[1];
> 21: TEX temp[1], input[0], 2D[2];
> 22: MUL output[0].xyz, temp[3], const[1].yyyy;
> 23: ADD temp[4], temp[4], const[1].zzzz;
> 24: MUL output[0].w, temp[1].xxxx, temp[4];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: TEX temp[0], input[0], 2D[0];
>  1: MOV temp[0].x, temp[0].wwww;
>  2: MAD temp[1], temp[0], const[2].xxxx, -const[2].yyyy;
>  3: MOV temp[1].z, const[2].zzzz;
>  4: DP3 temp[0].x, temp[1], temp[1];
>  5: ADD temp[0].x, const[2].yyyy, -temp[0].xxxx;
>  6: RSQ temp[0].x, temp[0].xxxx;
>  7: RCP temp[1].z, temp[0].xxxx;
>  8: DP3 temp[0].x, temp[1], input[2];
>  9: DP3 temp[0].y, temp[1], input[3];
> 10: DP3 temp[0].z, temp[1], input[4];
> 11: DP3 temp[2].w, input[1], input[1];
> 12: RSQ temp[2].w, temp[2].wwww;
> 13: MUL temp[2], input[1], temp[2].wwww;
> 14: DP3 temp[1], temp[2], temp[0];
> 15: MUL temp[1], temp[1], const[2].xxxx;
> 16: MAD temp[3], -temp[0], temp[1].xxxx, temp[2];
> 17: DP3_SAT temp[0], -temp[2], temp[0];
> 18: SUB temp[0], const[2].yyyy, temp[0];
> 19: POW temp[4], temp[0].xxxx, const[1].xxxx;
> 20: TEX temp[3], temp[3], CUBE[1];
> 21: TEX temp[1], input[0], 2D[2];
> 22: MUL output[0].xyz, temp[3], const[1].yyyy;
> 23: ADD temp[4], temp[4], const[1].zzzz;
> 24: MUL output[0].w, temp[1].xxxx, temp[4];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: TEX temp[0], input[0], 2D[0];
>  1: MOV temp[0].x, temp[0].wwww;
>  2: MAD temp[1], temp[0], const[2].xxxx, -const[2].yyyy;
>  3: MOV temp[1].z, const[2].zzzz;
>  4: DP3 temp[0].x, temp[1], temp[1];
>  5: ADD temp[0].x, const[2].yyyy, -temp[0].xxxx;
>  6: RSQ temp[0].x, |temp[0].xxxx|;
>  7: RCP temp[1].z, temp[0].xxxx;
>  8: DP3 temp[0].x, temp[1], input[2];
>  9: DP3 temp[0].y, temp[1], input[3];
> 10: DP3 temp[0].z, temp[1], input[4];
> 11: DP3 temp[2].w, input[1], input[1];
> 12: RSQ temp[2].w, |temp[2].wwww|;
> 13: MUL temp[2], input[1], temp[2].wwww;
> 14: DP3 temp[1], temp[2], temp[0];
> 15: MUL temp[1], temp[1], const[2].xxxx;
> 16: MAD temp[3], -temp[0], temp[1].xxxx, temp[2];
> 17: DP3_SAT temp[0], -temp[2], temp[0];
> 18: ADD temp[0], const[2].yyyy, -temp[0];
> 19: LG2 temp[5].w, temp[0].xxxx;
> 20: MUL temp[5].w, temp[5].wwww, const[1].xxxx;
> 21: EX2 temp[4], temp[5].wwww;
> 22: TEX temp[3], temp[3], CUBE[1];
> 23: TEX temp[1], input[0], 2D[2];
> 24: MUL output[0].xyz, temp[3], const[1].yyyy;
> 25: ADD temp[4], temp[4], const[1].zzzz;
> 26: MUL output[0].w, temp[1].xxxx, temp[4];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: TEX temp[0], input[0], 2D[0];
>  1: MOV temp[0].x, temp[0].wwww;
>  2: MAD temp[1], temp[0], const[2].xxxx, -const[2].yyyy;
>  3: MOV temp[1].z, const[2].zzzz;
>  4: DP3 temp[0].x, temp[1], temp[1];
>  5: ADD temp[0].x, const[2].yyyy, -temp[0].xxxx;
>  6: RSQ temp[0].x, |temp[0].xxxx|;
>  7: RCP temp[1].z, temp[0].xxxx;
>  8: DP3 temp[0].x, temp[1], input[2];
>  9: DP3 temp[0].y, temp[1], input[3];
> 10: DP3 temp[0].z, temp[1], input[4];
> 11: DP3 temp[2].w, input[1], input[1];
> 12: RSQ temp[2].w, |temp[2].wwww|;
> 13: MUL temp[2], input[1], temp[2].wwww;
> 14: DP3 temp[1], temp[2], temp[0];
> 15: MUL temp[1], temp[1], const[2].xxxx;
> 16: MAD temp[3], -temp[0], temp[1].xxxx, temp[2];
> 17: DP3_SAT temp[0], -temp[2], temp[0];
> 18: ADD temp[0], const[2].yyyy, -temp[0];
> 19: LG2 temp[5].w, temp[0].xxxx;
> 20: MUL temp[5].w, temp[5].wwww, const[1].xxxx;
> 21: EX2 temp[4], temp[5].wwww;
> 22: TEX temp[3], temp[3], CUBE[1];
> 23: TEX temp[1], input[0], 2D[2];
> 24: MUL output[0].xyz, temp[3], const[1].yyyy;
> 25: ADD temp[4], temp[4], const[1].zzzz;
> 26: MUL output[0].w, temp[1].xxxx, temp[4];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: TEX temp[0].yw, input[0].xy__, 2D[0];
>  1: MOV temp[0].x, temp[0].w___;
>  2: MAD temp[1].xy, temp[0].xy__, const[2].xx__, -const[2].yy__;
>  3: MOV temp[1].z, const[2].__z_;
>  4: DP3 temp[0].x, temp[1].xyz_, temp[1].xyz_;
>  5: ADD temp[0].x, const[2].y___, -temp[0].x___;
>  6: RSQ temp[0].x, |temp[0].x___|;
>  7: RCP temp[1].z, temp[0].x___;
>  8: DP3 temp[0].x, temp[1].xyz_, input[2].xyz_;
>  9: DP3 temp[0].y, temp[1].xyz_, input[3].xyz_;
> 10: DP3 temp[0].z, temp[1].xyz_, input[4].xyz_;
> 11: DP3 temp[2].w, input[1].xyz_, input[1].xyz_;
> 12: RSQ temp[2].w, |temp[2].w___|;
> 13: MUL temp[2].xyz, input[1].xyz_, temp[2].www_;
> 14: DP3 temp[1].x, temp[2].xyz_, temp[0].xyz_;
> 15: MUL temp[1].x, temp[1].x___, const[2].x___;
> 16: MAD temp[3].xyz, -temp[0].xyz_, temp[1].xxx_, temp[2].xyz_;
> 17: DP3_SAT temp[0].x, -temp[2].xyz_, temp[0].xyz_;
> 18: ADD temp[0].x, const[2].y___, -temp[0].x___;
> 19: LG2 temp[5].w, temp[0].x___;
> 20: MUL temp[5].w, temp[5].___w, const[1].___x;
> 21: EX2 temp[4].w, temp[5].w___;
> 22: TEX temp[3].xyz, temp[3].xyz_, CUBE[1];
> 23: TEX temp[1].x, input[0].xy__, 2D[2];
> 24: MUL output[0].xyz, temp[3].xyz_, const[1].yyy_;
> 25: ADD temp[4].w, temp[4].___w, const[1].___z;
> 26: MUL output[0].w, temp[1].___x, temp[4].___w;
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: TEX temp[0].yw, input[0].xy__, 2D[0];
>  1: MOV temp[0].x, temp[0].w___;
>  2: MAD temp[1].xy, temp[0].xy__, const[2].xx__, -none.11__;
>  3: MOV temp[1].z, none.__0_;
>  4: DP3 temp[0].x, temp[1].xyz_, temp[1].xyz_;
>  5: ADD temp[0].x, none.1___, -temp[0].x___;
>  6: RSQ temp[0].x, |temp[0].x___|;
>  7: RCP temp[1].z, temp[0].x___;
>  8: DP3 temp[0].x, temp[1].xyz_, input[2].xyz_;
>  9: DP3 temp[0].y, temp[1].xyz_, input[3].xyz_;
> 10: DP3 temp[0].z, temp[1].xyz_, input[4].xyz_;
> 11: DP3 temp[2].w, input[1].xyz_, input[1].xyz_;
> 12: RSQ temp[2].w, |temp[2].w___|;
> 13: MUL temp[2].xyz, input[1].xyz_, temp[2].www_;
> 14: DP3 temp[1].x, temp[2].xyz_, temp[0].xyz_;
> 15: MUL temp[1].x, temp[1].x___, const[2].x___;
> 16: MAD temp[3].xyz, -temp[0].xyz_, temp[1].xxx_, temp[2].xyz_;
> 17: DP3_SAT temp[0].x, -temp[2].xyz_, temp[0].xyz_;
> 18: ADD temp[0].x, none.1___, -temp[0].x___;
> 19: LG2 temp[5].w, temp[0].x___;
> 20: MUL temp[5].w, temp[5].___w, const[1].___x;
> 21: EX2 temp[4].w, temp[5].w___;
> 22: TEX temp[3].xyz, temp[3].xyz_, CUBE[1];
> 23: TEX temp[1].x, input[0].xy__, 2D[2];
> 24: MUL output[0].xyz, temp[3].xyz_, const[1].yyy_;
> 25: ADD temp[4].w, temp[4].___w, const[1].___z;
> 26: MUL output[0].w, temp[1].___x, temp[4].___w;
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: TEX temp[0].yw, input[0].xy__, 2D[0];
>  1: MOV temp[0].x, temp[0].w___;
>  2: MAD temp[1].xy, temp[0].xy__, const[2].xx__, -none.11__;
>  3: MOV temp[1].z, none.__0_;
>  4: DP3 temp[0].x, temp[1].xyz_, temp[1].xyz_;
>  5: ADD temp[0].x, none.1___, -temp[0].x___;
>  6: RSQ temp[0].x, |temp[0].x___|;
>  7: RCP temp[1].z, temp[0].x___;
>  8: DP3 temp[0].x, temp[1].xyz_, input[2].xyz_;
>  9: DP3 temp[0].y, temp[1].xyz_, input[3].xyz_;
> 10: DP3 temp[0].z, temp[1].xyz_, input[4].xyz_;
> 11: DP3 temp[2].w, input[1].xyz_, input[1].xyz_;
> 12: RSQ temp[2].w, |temp[2].w___|;
> 13: MUL temp[2].xyz, input[1].xyz_, temp[2].www_;
> 14: DP3 temp[1].x, temp[2].xyz_, temp[0].xyz_;
> 15: MUL temp[1].x, temp[1].x___, const[2].x___;
> 16: MAD temp[3].xyz, -temp[0].xyz_, temp[1].xxx_, temp[2].xyz_;
> 17: DP3_SAT temp[0].x, -temp[2].xyz_, temp[0].xyz_;
> 18: ADD temp[0].x, none.1___, -temp[0].x___;
> 19: LG2 temp[5].w, temp[0].x___;
> 20: MUL temp[5].w, temp[5].___w, const[1].___x;
> 21: EX2 temp[4].w, temp[5].w___;
> 22: TEX temp[3].xyz, temp[3].xyz_, CUBE[1];
> 23: TEX temp[1].x, input[0].xy__, 2D[2];
> 24: MUL output[0].xyz, temp[3].xyz_, const[1].yyy_;
> 25: ADD temp[4].w, temp[4].___w, const[1].___z;
> 26: MUL output[0].w, temp[1].___x, temp[4].___w;
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: TEX temp[0].yw, input[0].xy__, 2D[0];
>  1: src0.w = temp[0]
>     MAD temp[0].x, src0.w__, src0.111, src0.000
>  2: src0.xyz = temp[0], src1.xyz = const[2]
>     MAD temp[1].xy, src0.xy_, src1.xx_, -src0.11_
>  3: 
>     MAD temp[1].z, src0.__0, src0.111, src0.000
>  4: src0.xyz = temp[1]
>     DP3 temp[0].x, src0.xyz, src0.xyz
>  5: src0.xyz = temp[0]
>     MAD temp[0].x, src0.1__, src0.111, -src0.x__
>  6: src0.xyz = temp[0]
>     REPL_ALPHA temp[0].x
>     RSQ, |src0.x|
>  7: src0.xyz = temp[0]
>     REPL_ALPHA temp[1].z
>     RCP, src0.x
>  8: src0.xyz = temp[1], src1.xyz = input[2]
>     DP3 temp[0].x, src0.xyz, src1.xyz
>  9: src0.xyz = temp[1], src1.xyz = input[3]
>     DP3 temp[0].y, src0.xyz, src1.xyz
> 10: src0.xyz = temp[1], src1.xyz = input[4]
>     DP3 temp[0].z, src0.xyz, src1.xyz
> 11: src0.xyz = input[1]
>     DP3, src0.xyz, src0.xyz
>     DP3 temp[2].w, src0._, src0._
> 12: src0.w = temp[2]
>     RSQ temp[2].w, |src0.w|
> 13: src0.xyz = input[1], src0.w = temp[2]
>     MAD temp[2].xyz, src0.xyz, src0.www, src0.000
> 14: src0.xyz = temp[2], src1.xyz = temp[0]
>     DP3 temp[1].x, src0.xyz, src1.xyz
> 15: src0.xyz = temp[1], src1.xyz = const[2]
>     MAD temp[1].x, src0.x__, src1.x__, src0.000
> 16: src0.xyz = temp[0], src1.xyz = temp[1], src2.xyz = temp[2]
>     MAD temp[3].xyz, -src0.xyz, src1.xxx, src2.xyz
> 17: src0.xyz = temp[2], src1.xyz = temp[0]
>     DP3_SAT temp[0].x, -src0.xyz, src1.xyz
> 18: src0.xyz = temp[0]
>     MAD temp[0].x, src0.1__, src0.111, -src0.x__
> 19: src0.xyz = temp[0]
>     LG2 temp[5].w, src0.x
> 20: src0.xyz = const[1], src0.w = temp[5]
>     MAD temp[5].w, src0.w, src0.x, src0.0
> 21: src0.w = temp[5]
>     EX2 temp[4].w, src0.w
> 22: TEX temp[3].xyz, temp[3].xyz_, CUBE[1];
> 23: TEX temp[1].x, input[0].xy__, 2D[2];
> 24: src0.xyz = temp[3], src1.xyz = const[1]
>     MAD color[0].xyz, src0.xyz, src1.yyy, src0.000
> 25: src0.xyz = const[1], src0.w = temp[4]
>     MAD temp[4].w, src0.w, src0.1, src0.z
> 26: src0.xyz = temp[1], src0.w = temp[4]
>     MAD color[0].w, src0.x, src0.w, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[0].yw, input[0].xy__, 2D[0];
>  2: src0.xyz = input[1]
>     DP3, src0.xyz, src0.xyz
>     DP3 temp[2].w, src0._, src0._
>  3: src0.w = temp[0], src1.w = temp[2]
>     MAD temp[0].x, src0.w__, src0.111, src0.000
>     RSQ temp[2].w, |src1.w|
>  4: src0.xyz = input[1], src0.w = temp[2]
>     MAD temp[2].xyz, src0.xyz, src0.www, src0.000
>  5: src0.xyz = temp[0], src1.xyz = const[2]
>     MAD temp[1].xy, src0.xy_, src1.xx_, -src0.11_
>  6: 
>     MAD temp[1].z, src0.__0, src0.111, src0.000
>  7: src0.xyz = temp[1]
>     DP3 temp[0].x, src0.xyz, src0.xyz
>  8: src0.xyz = temp[0]
>     MAD temp[0].x, src0.1__, src0.111, -src0.x__
>  9: src0.xyz = temp[0]
>     REPL_ALPHA temp[0].x
>     RSQ, |src0.x|
> 10: src0.xyz = temp[0]
>     REPL_ALPHA temp[1].z
>     RCP, src0.x
> 11: src0.xyz = temp[1], src1.xyz = input[2]
>     DP3 temp[0].x, src0.xyz, src1.xyz
> 12: src0.xyz = temp[1], src1.xyz = input[3]
>     DP3 temp[0].y, src0.xyz, src1.xyz
> 13: src0.xyz = temp[1], src1.xyz = input[4]
>     DP3 temp[0].z, src0.xyz, src1.xyz
> 14: src0.xyz = temp[2], src1.xyz = temp[0]
>     DP3 temp[1].x, src0.xyz, src1.xyz
> 15: src0.xyz = temp[1], src1.xyz = const[2]
>     MAD temp[1].x, src0.x__, src1.x__, src0.000
> 16: src0.xyz = temp[0], src1.xyz = temp[1], src2.xyz = temp[2]
>     MAD temp[3].xyz, -src0.xyz, src1.xxx, src2.xyz
> 17: src0.xyz = temp[2], src1.xyz = temp[0]
>     DP3_SAT temp[0].x, -src0.xyz, src1.xyz
> 18: src0.xyz = temp[0]
>     MAD temp[0].x, src0.1__, src0.111, -src0.x__
> 19: src0.xyz = temp[0]
>     LG2 temp[5].w, src0.x
> 20: src0.xyz = const[1], src0.w = temp[5]
>     MAD temp[5].w, src0.w, src0.x, src0.0
> 21: src0.w = temp[5]
>     EX2 temp[4].w, src0.w
> 22: src0.xyz = const[1], src0.w = temp[4]
>     MAD temp[4].w, src0.w, src0.1, src0.z
> 23: BEGIN_TEX;
> 24: TEX temp[3].xyz, temp[3].xyz_, CUBE[1];
> 25: TEX temp[1].x, input[0].xy__, 2D[2];
> 26: src0.xyz = temp[3], src0.w = temp[4], src1.xyz = const[1], src2.xyz = temp[1]
>     MAD color[0].xyz, src0.xyz, src1.yyy, src0.000
>     MAD color[0].w, src2.x, src0.w, src0.0
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[5].yw, temp[0].xy__, 2D[0];
>  2: src0.xyz = temp[1]
>     DP3, src0.xyz, src0.xyz
>     DP3 temp[6].w, src0._, src0._
>  3: src0.w = temp[5], src1.w = temp[6]
>     MAD temp[5].x, src0.w__, src0.111, src0.000
>     RSQ temp[6].w, |src1.w|
>  4: src0.xyz = temp[1], src0.w = temp[6]
>     MAD temp[6].xyz, src0.xyz, src0.www, src0.000
>  5: src0.xyz = temp[5], src1.xyz = const[2]
>     MAD temp[1].xy, src0.xy_, src1.xx_, -src0.11_
>  6: 
>     MAD temp[1].z, src0.__0, src0.111, src0.000
>  7: src0.xyz = temp[1]
>     DP3 temp[5].x, src0.xyz, src0.xyz
>  8: src0.xyz = temp[5]
>     MAD temp[5].x, src0.1__, src0.111, -src0.x__
>  9: src0.xyz = temp[5]
>     REPL_ALPHA temp[5].x
>     RSQ, |src0.x|
> 10: src0.xyz = temp[5]
>     REPL_ALPHA temp[1].z
>     RCP, src0.x
> 11: src0.xyz = temp[1], src1.xyz = temp[2]
>     DP3 temp[5].x, src0.xyz, src1.xyz
> 12: src0.xyz = temp[1], src1.xyz = temp[3]
>     DP3 temp[5].y, src0.xyz, src1.xyz
> 13: src0.xyz = temp[1], src1.xyz = temp[4]
>     DP3 temp[5].z, src0.xyz, src1.xyz
> 14: src0.xyz = temp[6], src1.xyz = temp[5]
>     DP3 temp[1].x, src0.xyz, src1.xyz
> 15: src0.xyz = temp[1], src1.xyz = const[2]
>     MAD temp[1].x, src0.x__, src1.x__, src0.000
> 16: src0.xyz = temp[5], src1.xyz = temp[1], src2.xyz = temp[6]
>     MAD temp[2].xyz, -src0.xyz, src1.xxx, src2.xyz
> 17: src0.xyz = temp[6], src1.xyz = temp[5]
>     DP3_SAT temp[5].x, -src0.xyz, src1.xyz
> 18: src0.xyz = temp[5]
>     MAD temp[5].x, src0.1__, src0.111, -src0.x__
> 19: src0.xyz = temp[5]
>     LG2 temp[3].w, src0.x
> 20: src0.xyz = const[1], src0.w = temp[3]
>     MAD temp[3].w, src0.w, src0.x, src0.0
> 21: src0.w = temp[3]
>     EX2 temp[3].w, src0.w
> 22: src0.xyz = const[1], src0.w = temp[3]
>     MAD temp[3].w, src0.w, src0.1, src0.z
> 23: BEGIN_TEX;
> 24: TEX temp[2].xyz, temp[2].xyz_, CUBE[1];
> 25: TEX temp[1].x, temp[0].xy__, 2D[2];
> 26: src0.xyz = temp[2], src0.w = temp[3], src1.xyz = const[1], src2.xyz = temp[1]
>     MAD color[0].xyz, src0.xyz, src1.yyy, src0.000
>     MAD color[0].w, src2.x, src0.w, src0.0
>pc=43*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 20, tex_end: 0  (code_addr: 00000500)
>  TEX:
>    TEX t5, t0, texture[0]   (00008140)
>  0: xyz:  t1  t0  t0 ->                      (00000001)
>       w:  t0  t0  t0 -> t6.w                 (00980000)
>     xyz:   t1.xyz   t1.xyz   t1.xxx    op: 00804000
>       w:      1.0      1.0     t1.x    op: 00800891
>  1: xyz:  t0  t0  t0 -> t5.x                 (00940000)
>       w:  t5  t6  t0 -> t6.w                 (00980185)
>     xyz:   t5.www      1.0      0.0    op: 00050a8c
>       w:   |t6.w|     t0.x     t0.x    op: 0580004a
>  2: xyz:  t1  t0  t0 -> t6.xyz               (03980001)
>       w:  t6  t0  t0 ->                      (00000006)
>     xyz:   t1.xyz   t6.www      0.0    op: 00050600
>       w:     t1.x     t1.x     t1.x    op: 00000000
>  3: xyz:  t5  c2  t0 -> t1.xy                (01840885)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t5.xyz   c2.xxx     -1.0    op: 000d4280
>       w:     t5.x     t5.x     t5.x    op: 00000000
>  4: xyz:  t0  t0  t0 -> t1.z                 (02040000)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:      0.0      1.0      0.0    op: 00050a94
>       w:     t0.x     t0.x     t0.x    op: 00000000
>  5: xyz:  t1  t0  t0 -> t5.x                 (00940001)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   t1.xyz   t1.xxx    op: 00804000
>       w:     t1.x     t1.x     t1.x    op: 00000000
>  6: xyz:  t5  t0  t0 -> t5.x                 (00940005)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:      1.0      1.0  -t5.xyz    op: 00080a95
>       w:     t5.x     t5.x     t5.x    op: 00000000
>  7: xyz:  t5  t0  t0 -> t5.x                 (00940005)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t5.xxx   t5.xxx   t5.xxx    op: 05004081
>       w:   |t5.x|     t5.x     t5.x    op: 05800040
>  8: xyz:  t5  t0  t0 -> t1.z                 (02040005)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t5.xxx   t5.xxx   t5.xxx    op: 05004081
>       w:     t5.x     t5.x     t5.x    op: 05000000
>  9: xyz:  t1  t2  t0 -> t5.x                 (00940081)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   t2.xyz   t1.xxx    op: 00804200
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 10: xyz:  t1  t3  t0 -> t5.y                 (011400c1)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   t3.xyz   t1.xxx    op: 00804200
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 11: xyz:  t1  t4  t0 -> t5.z                 (02140101)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   t4.xyz   t1.xxx    op: 00804200
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 12: xyz:  t6  t5  t0 -> t1.x                 (00840146)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t6.xyz   t5.xyz   t6.xxx    op: 00804200
>       w:     t6.x     t6.x     t6.x    op: 00000000
> 13: xyz:  t1  c2  t0 -> t1.x                 (00840881)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   c2.xyz      0.0    op: 00050200
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 14: xyz:  t5  t1  t6 -> t2.xyz               (03886045)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:  -t5.xyz   t1.xxx   t6.xyz    op: 000202a0
>       w:     t5.x     t5.x     t5.x    op: 00000000
> 15: xyz:  t6  t5  t0 -> t5.x                 (00940146)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:  -t6.xyz   t5.xyz   t6.xxx    op: 40804220
>       w:     t6.x     t6.x     t6.x    op: 00000000
> 16: xyz:  t5  t0  t0 -> t5.x                 (00940005)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:      1.0      1.0  -t5.xyz    op: 00080a95
>       w:     t5.x     t5.x     t5.x    op: 00000000
> 17: xyz:  t5  t0  t0 ->                      (00000005)
>       w:  t0  t0  t0 -> t3.w                 (008c0000)
>     xyz:   t5.xxx   t5.xxx   t5.xxx    op: 00004081
>       w:     t5.x     t5.x     t5.x    op: 04800000
> 18: xyz:  c1  t0  t0 ->                      (00000021)
>       w:  t3  t0  t0 -> t3.w                 (008c0003)
>     xyz:   c1.xxx   c1.xxx   c1.xxx    op: 00004081
>       w:     t3.w     c1.x      0.0    op: 00040009
> 19: xyz:  t0  t0  t0 ->                      (00000000)
>       w:  t3  t0  t0 -> t3.w                 (008c0003)
>     xyz:   t0.xxx   t0.xxx   t0.xxx    op: 00004081
>       w:     t3.w     t0.x     t0.x    op: 04000009
> 20: xyz:  c1  t0  t0 ->                      (00000021)
>       w:  t3  t0  t0 -> t3.w                 (008c0003)
>     xyz:   c1.xxx   c1.xxx   c1.xxx    op: 00004081
>       w:     t3.w      1.0     c1.z    op: 00008889
>NODE 1: alu_offset: 21, tex_offset: 1, alu_end: 0, tex_end: 1  (code_addr: 00421015)
>  TEX:
>    TEX t2, t2, texture[1]   (00008882)
>    TEX t1, t0, texture[2]   (00009040)
> 21: xyz:  t2  c1  t1 -> o0.xyz               (1c001842)
>       w:  t3  t0  t0 -> o0.w                 (01000003)
>     xyz:   t2.xyz   c1.yyy      0.0    op: 00050300
>       w:     t1.x     t3.w      0.0    op: 00040486
>r300: Initial fragment program
>FRAG
>DCL IN[0], COLOR, LINEAR
>DCL IN[1], COLOR[1], LINEAR
>DCL IN[2], GENERIC[0], PERSPECTIVE
>DCL IN[3], GENERIC[1], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL SAMP[1]
>DCL SAMP[2]
>DCL SAMP[3]
>DCL CONST[2]
>DCL TEMP[0..1]
>IMM FLT32 {    8.0000,     4.0000,     2.0000,     0.0000}
>IMM FLT32 {    0.5000,     0.2500,     0.1250,     0.0625}
>IMM FLT32 {    1.0000,     0.0000,     0.0000,     0.0000}
>  0: MUL TEMP[0].xyz, IN[3], IMM[0].xxxx
>  1: TEX TEMP[0].x, TEMP[0], SAMP[0], 3D
>  2: MOV TEMP[0].w, TEMP[0].xxxx
>  3: MUL TEMP[1].xyz, IN[3], IMM[0].yyyy
>  4: TEX TEMP[1].x, TEMP[1], SAMP[0], 3D
>  5: MOV TEMP[0].z, TEMP[1].xxxx
>  6: MUL TEMP[1].xyz, IN[3], IMM[0].zzzz
>  7: TEX TEMP[1].x, TEMP[1], SAMP[0], 3D
>  8: MOV TEMP[0].y, TEMP[1].xxxx
>  9: TEX TEMP[0].x, IN[3], SAMP[0], 3D
> 10: DP4 TEMP[0].z, TEMP[0], IMM[1]
> 11: MAD TEMP[0].xy, TEMP[0].zzzz, CONST[2], IN[2]
> 12: TEX TEMP[0].x, TEMP[0], SAMP[2], 2D
> 13: MUL TEMP[0].x, TEMP[0], IMM[0].zzzz
> 14: ADD TEMP[0].y, TEMP[0].xxxx, -IMM[2].xxxx
> 15: MOV TEMP[0].w, IMM[0]
> 16: TEX TEMP[0].x, TEMP[0].zwzw, SAMP[1], 2D
> 17: ADD TEMP[0].x, TEMP[0], TEMP[0].yyyy
> 18: MOV TEMP[0].y, IMM[0].wwww
> 19: ADD TEMP[0].x, TEMP[0], -IN[0]
> 20: TEX TEMP[0], TEMP[0], SAMP[3], 2D
> 21: MAD TEMP[1].x, -IN[1], IN[1], IMM[2].xyyx
> 22: MUL TEMP[0], TEMP[0], IN[0].wwww
> 23: MUL OUT[0], TEMP[0], TEMP[1].xxxx
> 24: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: MUL temp[0].xyz, input[3], const[3].xxxx;
>  1: TEX temp[0].x, temp[0], 3D[0];
>  2: MOV temp[0].w, temp[0].xxxx;
>  3: MUL temp[1].xyz, input[3], const[3].yyyy;
>  4: TEX temp[1].x, temp[1], 3D[0];
>  5: MOV temp[0].z, temp[1].xxxx;
>  6: MUL temp[1].xyz, input[3], const[3].zzzz;
>  7: TEX temp[1].x, temp[1], 3D[0];
>  8: MOV temp[0].y, temp[1].xxxx;
>  9: TEX temp[0].x, input[3], 3D[0];
> 10: DP4 temp[0].z, temp[0], const[4];
> 11: MAD temp[0].xy, temp[0].zzzz, const[2], input[2];
> 12: TEX temp[0].x, temp[0], 2D[2];
> 13: MUL temp[0].x, temp[0], const[3].zzzz;
> 14: ADD temp[0].y, temp[0].xxxx, -temp[0].1111;
> 15: MOV temp[0].w, const[3];
> 16: TEX temp[0].x, temp[0].zwzw, 2D[1];
> 17: ADD temp[0].x, temp[0], temp[0].yyyy;
> 18: MOV temp[0].y, const[3].wwww;
> 19: ADD temp[0].x, temp[0], -input[0];
> 20: TEX temp[0], temp[0], 2D[3];
> 21: MAD temp[1].x, -input[1], input[1], temp[0].1001;
> 22: MUL temp[0], temp[0], input[0].wwww;
> 23: MUL output[0], temp[0], temp[1].xxxx;
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: MUL temp[0].xyz, input[3], const[3].xxxx;
>  1: TEX temp[0].x, temp[0], 3D[0];
>  2: MOV temp[0].w, temp[0].xxxx;
>  3: MUL temp[1].xyz, input[3], const[3].yyyy;
>  4: TEX temp[1].x, temp[1], 3D[0];
>  5: MOV temp[0].z, temp[1].xxxx;
>  6: MUL temp[1].xyz, input[3], const[3].zzzz;
>  7: TEX temp[1].x, temp[1], 3D[0];
>  8: MOV temp[0].y, temp[1].xxxx;
>  9: TEX temp[0].x, input[3], 3D[0];
> 10: DP4 temp[0].z, temp[0], const[4];
> 11: MAD temp[0].xy, temp[0].zzzz, const[2], input[2];
> 12: TEX temp[0].x, temp[0], 2D[2];
> 13: MUL temp[0].x, temp[0], const[3].zzzz;
> 14: ADD temp[0].y, temp[0].xxxx, -temp[0].1111;
> 15: MOV temp[0].w, const[3];
> 16: TEX temp[0].x, temp[0].zwzw, 2D[1];
> 17: ADD temp[0].x, temp[0], temp[0].yyyy;
> 18: MOV temp[0].y, const[3].wwww;
> 19: ADD temp[0].x, temp[0], -input[0];
> 20: TEX temp[0], temp[0], 2D[3];
> 21: MAD temp[1].x, -input[1], input[1], temp[0].1001;
> 22: MUL temp[0], temp[0], input[0].wwww;
> 23: MUL output[0], temp[0], temp[1].xxxx;
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: MUL temp[0].xyz, input[3], const[3].xxxx;
>  1: TEX temp[0].x, temp[0], 3D[0];
>  2: MOV temp[0].w, temp[0].xxxx;
>  3: MUL temp[1].xyz, input[3], const[3].yyyy;
>  4: TEX temp[1].x, temp[1], 3D[0];
>  5: MOV temp[0].z, temp[1].xxxx;
>  6: MUL temp[1].xyz, input[3], const[3].zzzz;
>  7: TEX temp[1].x, temp[1], 3D[0];
>  8: MOV temp[0].y, temp[1].xxxx;
>  9: TEX temp[0].x, input[3], 3D[0];
> 10: DP4 temp[0].z, temp[0], const[4];
> 11: MAD temp[0].xy, temp[0].zzzz, const[2], input[2];
> 12: TEX temp[0].x, temp[0], 2D[2];
> 13: MUL temp[0].x, temp[0], const[3].zzzz;
> 14: ADD temp[0].y, temp[0].xxxx, -temp[0].1111;
> 15: MOV temp[0].w, const[3];
> 16: TEX temp[0].x, temp[0].zwzw, 2D[1];
> 17: ADD temp[0].x, temp[0], temp[0].yyyy;
> 18: MOV temp[0].y, const[3].wwww;
> 19: ADD temp[0].x, temp[0], -input[0];
> 20: TEX temp[0], temp[0], 2D[3];
> 21: MAD temp[1].x, -input[1], input[1], temp[0].1001;
> 22: MUL temp[0], temp[0], input[0].wwww;
> 23: MUL output[0], temp[0], temp[1].xxxx;
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: MUL temp[0].xyz, input[3], const[3].xxxx;
>  1: TEX temp[0].x, temp[0], 3D[0];
>  2: MOV temp[0].w, temp[0].xxxx;
>  3: MUL temp[1].xyz, input[3], const[3].yyyy;
>  4: TEX temp[1].x, temp[1], 3D[0];
>  5: MOV temp[0].z, temp[1].xxxx;
>  6: MUL temp[1].xyz, input[3], const[3].zzzz;
>  7: TEX temp[1].x, temp[1], 3D[0];
>  8: MOV temp[0].y, temp[1].xxxx;
>  9: TEX temp[0].x, input[3], 3D[0];
> 10: DP4 temp[0].z, temp[0], const[4];
> 11: MAD temp[0].xy, temp[0].zzzz, const[2], input[2];
> 12: TEX temp[0].x, temp[0], 2D[2];
> 13: MUL temp[0].x, temp[0], const[3].zzzz;
> 14: ADD temp[0].y, temp[0].xxxx, -temp[0].1111;
> 15: MOV temp[0].w, const[3];
> 16: TEX temp[0].x, temp[0].zwzw, 2D[1];
> 17: ADD temp[0].x, temp[0], temp[0].yyyy;
> 18: MOV temp[0].y, const[3].wwww;
> 19: ADD temp[0].x, temp[0], -input[0];
> 20: TEX temp[0], temp[0], 2D[3];
> 21: MAD temp[1].x, -input[1], input[1], temp[0].1001;
> 22: MUL temp[0], temp[0], input[0].wwww;
> 23: MUL output[0], temp[0], temp[1].xxxx;
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: MUL temp[0].xyz, input[3], const[3].xxxx;
>  1: TEX temp[2], temp[0], 3D[0];
>  2: MOV temp[0].x, temp[2];
>  3: MOV temp[0].w, temp[0].xxxx;
>  4: MUL temp[1].xyz, input[3], const[3].yyyy;
>  5: TEX temp[3], temp[1], 3D[0];
>  6: MOV temp[1].x, temp[3];
>  7: MOV temp[0].z, temp[1].xxxx;
>  8: MUL temp[1].xyz, input[3], const[3].zzzz;
>  9: TEX temp[4], temp[1], 3D[0];
> 10: MOV temp[1].x, temp[4];
> 11: MOV temp[0].y, temp[1].xxxx;
> 12: TEX temp[5], input[3], 3D[0];
> 13: MOV temp[0].x, temp[5];
> 14: DP4 temp[0].z, temp[0], const[4];
> 15: MAD temp[0].xy, temp[0].zzzz, const[2], input[2];
> 16: TEX temp[6], temp[0], 2D[2];
> 17: MOV temp[0].x, temp[6];
> 18: MUL temp[0].x, temp[0], const[3].zzzz;
> 19: ADD temp[0].y, temp[0].xxxx, -temp[0].1111;
> 20: MOV temp[0].w, const[3];
> 21: TEX temp[7], temp[0].zwzw, 2D[1];
> 22: MOV temp[0].x, temp[7];
> 23: ADD temp[0].x, temp[0], temp[0].yyyy;
> 24: MOV temp[0].y, const[3].wwww;
> 25: ADD temp[0].x, temp[0], -input[0];
> 26: TEX temp[0], temp[0], 2D[3];
> 27: MAD temp[1].x, -input[1], input[1], temp[0].1001;
> 28: MUL temp[0], temp[0], input[0].wwww;
> 29: MUL output[0], temp[0], temp[1].xxxx;
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: MUL temp[0].xyz, input[3].xyz_, const[3].xxx_;
>  1: TEX temp[2].x, temp[0].xyz_, 3D[0];
>  2: MOV temp[0].x, temp[2].x___;
>  3: MOV temp[0].w, temp[0].___x;
>  4: MUL temp[1].xyz, input[3].xyz_, const[3].yyy_;
>  5: TEX temp[3].x, temp[1].xyz_, 3D[0];
>  6: MOV temp[1].x, temp[3].x___;
>  7: MOV temp[0].z, temp[1].__x_;
>  8: MUL temp[1].xyz, input[3].xyz_, const[3].zzz_;
>  9: TEX temp[4].x, temp[1].xyz_, 3D[0];
> 10: MOV temp[1].x, temp[4].x___;
> 11: MOV temp[0].y, temp[1]._x__;
> 12: TEX temp[5].x, input[3].xyz_, 3D[0];
> 13: MOV temp[0].x, temp[5].x___;
> 14: DP4 temp[0].z, temp[0], const[4];
> 15: MAD temp[0].xy, temp[0].zz__, const[2].xy__, input[2].xy__;
> 16: TEX temp[6].x, temp[0].xy__, 2D[2];
> 17: MOV temp[0].x, temp[6].x___;
> 18: MUL temp[0].x, temp[0].x___, const[3].z___;
> 19: ADD temp[0].y, temp[0]._x__, -temp[0]._1__;
> 20: MOV temp[0].w, const[3].___w;
> 21: TEX temp[7].x, temp[0].zw__, 2D[1];
> 22: MOV temp[0].x, temp[7].x___;
> 23: ADD temp[0].x, temp[0].x___, temp[0].y___;
> 24: MOV temp[0].y, const[3]._w__;
> 25: ADD temp[0].x, temp[0].x___, -input[0].x___;
> 26: TEX temp[0], temp[0].xy__, 2D[3];
> 27: MAD temp[1].x, -input[1].x___, input[1].x___, temp[0].1___;
> 28: MUL temp[0], temp[0], input[0].wwww;
> 29: MUL output[0], temp[0], temp[1].xxxx;
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: MUL temp[0].xyz, input[3].xyz_, const[3].xxx_;
>  1: TEX temp[2].x, temp[0].xyz_, 3D[0];
>  2: MOV temp[0].w, temp[2].___x;
>  3: MUL temp[1].xyz, input[3].xyz_, const[3].yyy_;
>  4: TEX temp[3].x, temp[1].xyz_, 3D[0];
>  5: MOV temp[0].z, temp[3].__x_;
>  6: MUL temp[1].xyz, input[3].xyz_, const[3].zzz_;
>  7: TEX temp[4].x, temp[1].xyz_, 3D[0];
>  8: MOV temp[0].y, temp[4]._x__;
>  9: TEX temp[5].x, input[3].xyz_, 3D[0];
> 10: MOV temp[0].x, temp[5].x___;
> 11: DP4 temp[0].z, temp[0], const[4];
> 12: MAD temp[0].xy, temp[0].zz__, const[2].xy__, input[2].xy__;
> 13: TEX temp[6].x, temp[0].xy__, 2D[2];
> 14: MUL temp[0].x, temp[6].x___, const[3].z___;
> 15: ADD temp[0].y, temp[0]._x__, -temp[2]._1__;
> 16: MOV temp[0].w, none.___0;
> 17: TEX temp[7].x, temp[0].zw__, 2D[1];
> 18: ADD temp[0].x, temp[7].x___, temp[0].y___;
> 19: MOV temp[0].y, none._0__;
> 20: ADD temp[0].x, temp[0].x___, -input[0].x___;
> 21: TEX temp[0], temp[0].xy__, 2D[3];
> 22: MAD temp[1].x, -input[1].x___, input[1].x___, temp[2].1___;
> 23: MUL temp[0], temp[0], input[0].wwww;
> 24: MUL output[0], temp[0], temp[1].xxxx;
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: MUL temp[0].xyz, input[3].xyz_, const[3].xxx_;
>  1: TEX temp[2].x, temp[0].xyz_, 3D[0];
>  2: MOV temp[0].w, temp[2].___x;
>  3: MUL temp[1].xyz, input[3].xyz_, const[3].yyy_;
>  4: TEX temp[3].x, temp[1].xyz_, 3D[0];
>  5: MOV temp[0].z, temp[3].__x_;
>  6: MUL temp[1].xyz, input[3].xyz_, const[3].zzz_;
>  7: TEX temp[4].x, temp[1].xyz_, 3D[0];
>  8: MOV temp[0].y, temp[4]._x__;
>  9: TEX temp[5].x, input[3].xyz_, 3D[0];
> 10: MOV temp[0].x, temp[5].x___;
> 11: DP4 temp[0].z, temp[0], const[4];
> 12: MAD temp[0].xy, temp[0].zz__, const[2].xy__, input[2].xy__;
> 13: TEX temp[6].x, temp[0].xy__, 2D[2];
> 14: MUL temp[0].x, temp[6].x___, const[3].z___;
> 15: ADD temp[0].y, temp[0]._x__, -temp[2]._1__;
> 16: MOV temp[0].w, none.___0;
> 17: MOV temp[8].x, temp[0].z___;
> 18: MOV temp[8].y, temp[0]._w__;
> 19: TEX temp[7].x, temp[8].xy__, 2D[1];
> 20: ADD temp[0].x, temp[7].x___, temp[0].y___;
> 21: MOV temp[0].y, none._0__;
> 22: ADD temp[0].x, temp[0].x___, -input[0].x___;
> 23: TEX temp[0], temp[0].xy__, 2D[3];
> 24: MAD temp[1].x, -input[1].x___, input[1].x___, temp[2].1___;
> 25: MUL temp[0], temp[0], input[0].wwww;
> 26: MUL output[0], temp[0], temp[1].xxxx;
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: src0.xyz = input[3], src1.xyz = const[3]
>     MAD temp[0].xyz, src0.xyz, src1.xxx, src0.000
>  1: TEX temp[2].x, temp[0].xyz_, 3D[0];
>  2: src0.xyz = temp[2]
>     MAD temp[0].w, src0.x, src0.1, src0.0
>  3: src0.xyz = input[3], src1.xyz = const[3]
>     MAD temp[1].xyz, src0.xyz, src1.yyy, src0.000
>  4: TEX temp[3].x, temp[1].xyz_, 3D[0];
>  5: src0.xyz = temp[3]
>     MAD temp[0].z, src0.__x, src0.111, src0.000
>  6: src0.xyz = input[3], src1.xyz = const[3]
>     MAD temp[1].xyz, src0.xyz, src1.zzz, src0.000
>  7: TEX temp[4].x, temp[1].xyz_, 3D[0];
>  8: src0.xyz = temp[4]
>     MAD temp[0].y, src0._x_, src0.111, src0.000
>  9: TEX temp[5].x, input[3].xyz_, 3D[0];
> 10: src0.xyz = temp[5]
>     MAD temp[0].x, src0.x__, src0.111, src0.000
> 11: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[4], src1.w = const[4]
>     DP4 temp[0].z, src0.xyz, src1.xyz
>     DP4, src0.w, src1.w
> 12: src0.xyz = temp[0], src1.xyz = const[2], src2.xyz = input[2]
>     MAD temp[0].xy, src0.zz_, src1.xy_, src2.xy_
> 13: TEX temp[6].x, temp[0].xy__, 2D[2];
> 14: src0.xyz = temp[6], src1.xyz = const[3]
>     MAD temp[0].x, src0.x__, src1.z__, src0.000
> 15: src0.xyz = temp[0]
>     MAD temp[0].y, src0._x_, src0.111, -src0._1_
> 16: 
>     MAD temp[0].w, src0.0, src0.1, src0.0
> 17: src0.xyz = temp[0]
>     MAD temp[8].x, src0.z__, src0.111, src0.000
> 18: src0.w = temp[0]
>     MAD temp[8].y, src0._w_, src0.111, src0.000
> 19: TEX temp[7].x, temp[8].xy__, 2D[1];
> 20: src0.xyz = temp[7], src1.xyz = temp[0]
>     MAD temp[0].x, src0.x__, src0.111, src1.y__
> 21: 
>     MAD temp[0].y, src0._0_, src0.111, src0.000
> 22: src0.xyz = temp[0], src1.xyz = input[0]
>     MAD temp[0].x, src0.x__, src0.111, -src1.x__
> 23: TEX temp[0], temp[0].xy__, 2D[3];
> 24: src0.xyz = input[1]
>     MAD temp[1].x, -src0.x__, src0.x__, src0.1__
> 25: src0.xyz = temp[0], src0.w = temp[0], src1.w = input[0]
>     MAD temp[0].xyz, src0.xyz, src1.www, src0.000
>     MAD temp[0].w, src0.w, src1.w, src0.0
> 26: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = temp[1]
>     MAD color[0].xyz, src0.xyz, src1.xxx, src0.000
>     MAD color[0].w, src0.w, src1.x, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[5].x, input[3].xyz_, 3D[0];
>  2: src0.xyz = input[3], src1.xyz = const[3]
>     MAD temp[1].xyz, src0.xyz, src1.yyy, src0.000
>  3: src0.xyz = input[3], src1.xyz = const[3]
>     MAD temp[0].xyz, src0.xyz, src1.xxx, src0.000
>  4: BEGIN_TEX;
>  5: TEX temp[2].x, temp[0].xyz_, 3D[0];
>  6: TEX temp[3].x, temp[1].xyz_, 3D[0];
>  7: src0.xyz = temp[3], src1.xyz = temp[2]
>     MAD temp[0].z, src0.__x, src0.111, src0.000
>     MAD temp[0].w, src1.x, src0.1, src0.0
>  8: src0.xyz = input[3], src1.xyz = const[3]
>     MAD temp[1].xyz, src0.xyz, src1.zzz, src0.000
>  9: src0.xyz = temp[5]
>     MAD temp[0].x, src0.x__, src0.111, src0.000
> 10: BEGIN_TEX;
> 11: TEX temp[4].x, temp[1].xyz_, 3D[0];
> 12: src0.xyz = temp[4]
>     MAD temp[0].y, src0._x_, src0.111, src0.000
> 13: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = const[4], src1.w = const[4]
>     DP4 temp[0].z, src0.xyz, src1.xyz
>     DP4, src0.w, src1.w
> 14: src0.xyz = temp[0], src1.xyz = const[2], src2.xyz = input[2]
>     MAD temp[0].xy, src0.zz_, src1.xy_, src2.xy_
>     MAD temp[0].w, src0.0, src0.1, src0.0
> 15: src0.w = temp[0]
>     MAD temp[8].y, src0._w_, src0.111, src0.000
> 16: src0.xyz = temp[0]
>     MAD temp[8].x, src0.z__, src0.111, src0.000
> 17: src0.xyz = input[1]
>     MAD temp[1].x, -src0.x__, src0.x__, src0.1__
> 18: BEGIN_TEX;
> 19: TEX temp[7].x, temp[8].xy__, 2D[1];
> 20: TEX temp[6].x, temp[0].xy__, 2D[2];
> 21: src0.xyz = temp[6], src1.xyz = const[3]
>     MAD temp[0].x, src0.x__, src1.z__, src0.000
> 22: src0.xyz = temp[0]
>     MAD temp[0].y, src0._x_, src0.111, -src0._1_
> 23: src0.xyz = temp[7], src1.xyz = temp[0]
>     MAD temp[0].x, src0.x__, src0.111, src1.y__
> 24: src0.xyz = temp[0], src1.xyz = input[0]
>     MAD temp[0].x, src0.x__, src0.111, -src1.x__
> 25: 
>     MAD temp[0].y, src0._0_, src0.111, src0.000
> 26: BEGIN_TEX;
> 27: TEX temp[0], temp[0].xy__, 2D[3];
> 28: src0.xyz = temp[0], src0.w = temp[0], src1.w = input[0]
>     MAD temp[0].xyz, src0.xyz, src1.www, src0.000
>     MAD temp[0].w, src0.w, src1.w, src0.0
> 29: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = temp[1]
>     MAD color[0].xyz, src0.xyz, src1.xxx, src0.000
>     MAD color[0].w, src0.w, src1.x, src0.0
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[8].x, temp[3].xyz_, 3D[0];
>  2: src0.xyz = temp[3], src1.xyz = const[3]
>     MAD temp[5].xyz, src0.xyz, src1.yyy, src0.000
>  3: src0.xyz = temp[3], src1.xyz = const[3]
>     MAD temp[4].xyz, src0.xyz, src1.xxx, src0.000
>  4: BEGIN_TEX;
>  5: TEX temp[6].x, temp[4].xyz_, 3D[0];
>  6: TEX temp[7].x, temp[5].xyz_, 3D[0];
>  7: src0.xyz = temp[7], src1.xyz = temp[6]
>     MAD temp[4].z, src0.__x, src0.111, src0.000
>     MAD temp[4].w, src1.x, src0.1, src0.0
>  8: src0.xyz = temp[3], src1.xyz = const[3]
>     MAD temp[5].xyz, src0.xyz, src1.zzz, src0.000
>  9: src0.xyz = temp[8]
>     MAD temp[4].x, src0.x__, src0.111, src0.000
> 10: BEGIN_TEX;
> 11: TEX temp[3].x, temp[5].xyz_, 3D[0];
> 12: src0.xyz = temp[3]
>     MAD temp[4].y, src0._x_, src0.111, src0.000
> 13: src0.xyz = temp[4], src0.w = temp[4], src1.xyz = const[4], src1.w = const[4]
>     DP4 temp[4].z, src0.xyz, src1.xyz
>     DP4, src0.w, src1.w
> 14: src0.xyz = temp[4], src1.xyz = const[2], src2.xyz = temp[2]
>     MAD temp[4].xy, src0.zz_, src1.xy_, src2.xy_
>     MAD temp[4].w, src0.0, src0.1, src0.0
> 15: src0.w = temp[4]
>     MAD temp[2].y, src0._w_, src0.111, src0.000
> 16: src0.xyz = temp[4]
>     MAD temp[2].x, src0.z__, src0.111, src0.000
> 17: src0.xyz = temp[1]
>     MAD temp[5].x, -src0.x__, src0.x__, src0.1__
> 18: BEGIN_TEX;
> 19: TEX temp[2].x, temp[2].xy__, 2D[1];
> 20: TEX temp[1].x, temp[4].xy__, 2D[2];
> 21: src0.xyz = temp[1], src1.xyz = const[3]
>     MAD temp[4].x, src0.x__, src1.z__, src0.000
> 22: src0.xyz = temp[4]
>     MAD temp[4].y, src0._x_, src0.111, -src0._1_
> 23: src0.xyz = temp[2], src1.xyz = temp[4]
>     MAD temp[4].x, src0.x__, src0.111, src1.y__
> 24: src0.xyz = temp[4], src1.xyz = temp[0]
>     MAD temp[4].x, src0.x__, src0.111, -src1.x__
> 25: 
>     MAD temp[4].y, src0._0_, src0.111, src0.000
> 26: BEGIN_TEX;
> 27: TEX temp[4], temp[4].xy__, 2D[3];
> 28: src0.xyz = temp[4], src0.w = temp[4], src1.w = temp[0]
>     MAD temp[4].xyz, src0.xyz, src1.www, src0.000
>     MAD temp[4].w, src0.w, src1.w, src0.0
> 29: src0.xyz = temp[4], src0.w = temp[4], src1.xyz = temp[5]
>     MAD color[0].xyz, src0.xyz, src1.xxx, src0.000
>     MAD color[0].w, src0.w, src1.x, src0.0
>r300compiler error: r300_fragprog_emit.c::begin_tex(): Too many texture indirections
>pc=44*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 0, tex_end: 0  (code_addr: 00000000)
>  TEX:
>    TEX t8, t3, texture[0]   (00008203)
>  0: xyz:  t3  c3  t0 -> t5.xyz               (039408c3)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t3.xyz   c3.yyy      0.0    op: 00050300
>       w:     t3.x     t3.x     t3.x    op: 00000000
>r300 FP: Compiler Error:
>r300_fragprog_emit.c::begin_tex(): Too many texture indirections
>Using a dummy shader instead.
>If there's an 'unknown opcode' message, please file a bug report and attach this log.
>r300: Initial fragment program
>FRAG
>DCL OUT[0], COLOR
>IMM FLT32 {    0.0000,     1.0000,     0.0000,     0.0000}
>  0: MOV OUT[0], IMM[0].xxxy
>  1: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0001;
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0001;
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0001;
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0001;
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0001;
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0001;
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0001;
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: MOV output[0], temp[0].0001;
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: 
>     MAD color[0].xyz, src0.000, src0.111, src0.000
>     MAD color[0].w, src0.1, src0.1, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: 
>     MAD color[0].xyz, src0.000, src0.111, src0.000
>     MAD color[0].w, src0.1, src0.1, src0.0
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: 
>     MAD color[0].xyz, src0.000, src0.111, src0.000
>     MAD color[0].w, src0.1, src0.1, src0.0
>pc=45*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 0, tex_end: 0  (code_addr: 00400000)
>  0: xyz:  t0  t0  t0 -> o0.xyz               (1c000000)
>       w:  t0  t0  t0 -> o0.w                 (01000000)
>     xyz:      0.0      1.0      0.0    op: 00050a94
>       w:      1.0      1.0      0.0    op: 00040891
>r300: Initial fragment program
>FRAG
>DCL IN[0], COLOR, LINEAR
>DCL IN[1], GENERIC[0], PERSPECTIVE
>DCL IN[2], GENERIC[1], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL SAMP[1]
>DCL TEMP[0..1]
>  0: TEX TEMP[0], IN[2], SAMP[1], 2D
>  1: TEX TEMP[1], IN[1], SAMP[0], 2D
>  2: MUL TEMP[1], TEMP[1], TEMP[0]
>  3: MUL OUT[0], TEMP[1], IN[0]
>  4: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: TEX temp[0], input[2], 2D[1];
>  1: TEX temp[1], input[1], 2D[0];
>  2: MUL temp[1], temp[1], temp[0];
>  3: MUL output[0], temp[1], input[0];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: TEX temp[0], input[2], 2D[1];
>  1: TEX temp[1], input[1], 2D[0];
>  2: MUL temp[1], temp[1], temp[0];
>  3: MUL output[0], temp[1], input[0];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: TEX temp[0], input[2], 2D[1];
>  1: TEX temp[1], input[1], 2D[0];
>  2: MUL temp[1], temp[1], temp[0];
>  3: MUL output[0], temp[1], input[0];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: TEX temp[0], input[2], 2D[1];
>  1: TEX temp[1], input[1], 2D[0];
>  2: MUL temp[1], temp[1], temp[0];
>  3: MUL output[0], temp[1], input[0];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: TEX temp[0], input[2], 2D[1];
>  1: TEX temp[1], input[1], 2D[0];
>  2: MUL temp[1], temp[1], temp[0];
>  3: MUL output[0], temp[1], input[0];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: TEX temp[0], input[2].xy__, 2D[1];
>  1: TEX temp[1], input[1].xy__, 2D[0];
>  2: MUL temp[1], temp[1], temp[0];
>  3: MUL output[0], temp[1], input[0];
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: TEX temp[0], input[2].xy__, 2D[1];
>  1: TEX temp[1], input[1].xy__, 2D[0];
>  2: MUL temp[1], temp[1], temp[0];
>  3: MUL output[0], temp[1], input[0];
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: TEX temp[0], input[2].xy__, 2D[1];
>  1: TEX temp[1], input[1].xy__, 2D[0];
>  2: MUL temp[1], temp[1], temp[0];
>  3: MUL output[0], temp[1], input[0];
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: TEX temp[0], input[2].xy__, 2D[1];
>  1: TEX temp[1], input[1].xy__, 2D[0];
>  2: src0.xyz = temp[1], src0.w = temp[1], src1.xyz = temp[0], src1.w = temp[0]
>     MAD temp[1].xyz, src0.xyz, src1.xyz, src0.000
>     MAD temp[1].w, src0.w, src1.w, src0.0
>  3: src0.xyz = temp[1], src0.w = temp[1], src1.xyz = input[0], src1.w = input[0]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[1], input[1].xy__, 2D[0];
>  2: TEX temp[0], input[2].xy__, 2D[1];
>  3: src0.xyz = temp[1], src0.w = temp[1], src1.xyz = temp[0], src1.w = temp[0]
>     MAD temp[1].xyz, src0.xyz, src1.xyz, src0.000
>     MAD temp[1].w, src0.w, src1.w, src0.0
>  4: src0.xyz = temp[1], src0.w = temp[1], src1.xyz = input[0], src1.w = input[0]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[3], temp[1].xy__, 2D[0];
>  2: TEX temp[1], temp[2].xy__, 2D[1];
>  3: src0.xyz = temp[3], src0.w = temp[3], src1.xyz = temp[1], src1.w = temp[1]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src0.000
>     MAD temp[3].w, src0.w, src1.w, src0.0
>  4: src0.xyz = temp[3], src0.w = temp[3], src1.xyz = temp[0], src1.w = temp[0]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>pc=46*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 1, tex_end: 1  (code_addr: 00420040)
>  TEX:
>    TEX t3, t1, texture[0]   (000080c1)
>    TEX t1, t2, texture[1]   (00008842)
>  0: xyz:  t3  t1  t0 -> t3.xyz               (038c0043)
>       w:  t3  t1  t0 -> t3.w                 (008c0043)
>     xyz:   t3.xyz   t1.xyz      0.0    op: 00050200
>       w:     t3.w     t1.w      0.0    op: 00040509
>  1: xyz:  t3  t0  t0 -> o0.xyz               (1c000003)
>       w:  t3  t0  t0 -> o0.w                 (01000003)
>     xyz:   t3.xyz   t0.xyz      0.0    op: 00050200
>       w:     t3.w     t0.w      0.0    op: 00040509
>r300: Initial fragment program
>FRAG
>DCL IN[0], GENERIC[2], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL CONST[0]
>DCL TEMP[0]
>IMM FLT32 {    2.7183,     1.0000,     0.0000,     0.0000}
>  0: DP3 TEMP[0].x, IN[0], IN[0]
>  1: RSQ TEMP[0].x, TEMP[0].xxxx
>  2: RCP TEMP[0].x, TEMP[0].xxxx
>  3: MUL TEMP[0].x, TEMP[0], CONST[0].wwww
>  4: MUL TEMP[0].x, TEMP[0], TEMP[0]
>  5: POW TEMP[0].x, IMM[0].xxxx, -TEMP[0].xxxx
>  6: MOV OUT[0].xyz, CONST[0]
>  7: ADD OUT[0].w, -TEMP[0].xxxx, IMM[0].yyyy
>  8: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: DP3 temp[0].x, input[0], input[0];
>  1: RSQ temp[0].x, temp[0].xxxx;
>  2: RCP temp[0].x, temp[0].xxxx;
>  3: MUL temp[0].x, temp[0], const[0].wwww;
>  4: MUL temp[0].x, temp[0], temp[0];
>  5: POW temp[0].x, const[1].xxxx, -temp[0].xxxx;
>  6: MOV output[0].xyz, const[0];
>  7: ADD output[0].w, -temp[0].xxxx, const[1].yyyy;
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: DP3 temp[0].x, input[0], input[0];
>  1: RSQ temp[0].x, temp[0].xxxx;
>  2: RCP temp[0].x, temp[0].xxxx;
>  3: MUL temp[0].x, temp[0], const[0].wwww;
>  4: MUL temp[0].x, temp[0], temp[0];
>  5: POW temp[0].x, const[1].xxxx, -temp[0].xxxx;
>  6: MOV output[0].xyz, const[0];
>  7: ADD output[0].w, -temp[0].xxxx, const[1].yyyy;
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: DP3 temp[0].x, input[0], input[0];
>  1: RSQ temp[0].x, temp[0].xxxx;
>  2: RCP temp[0].x, temp[0].xxxx;
>  3: MUL temp[0].x, temp[0], const[0].wwww;
>  4: MUL temp[0].x, temp[0], temp[0];
>  5: POW temp[0].x, const[1].xxxx, -temp[0].xxxx;
>  6: MOV output[0].xyz, const[0];
>  7: ADD output[0].w, -temp[0].xxxx, const[1].yyyy;
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: DP3 temp[0].x, input[0], input[0];
>  1: RSQ temp[0].x, |temp[0].xxxx|;
>  2: RCP temp[0].x, temp[0].xxxx;
>  3: MUL temp[0].x, temp[0], const[0].wwww;
>  4: MUL temp[0].x, temp[0], temp[0];
>  5: LG2 temp[1].w, const[1].xxxx;
>  6: MUL temp[1].w, temp[1].wwww, -temp[0].xxxx;
>  7: EX2 temp[0].x, temp[1].wwww;
>  8: MOV output[0].xyz, const[0];
>  9: ADD output[0].w, -temp[0].xxxx, const[1].yyyy;
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: DP3 temp[0].x, input[0], input[0];
>  1: RSQ temp[0].x, |temp[0].xxxx|;
>  2: RCP temp[0].x, temp[0].xxxx;
>  3: MUL temp[0].x, temp[0], const[0].wwww;
>  4: MUL temp[0].x, temp[0], temp[0];
>  5: LG2 temp[1].w, const[1].xxxx;
>  6: MUL temp[1].w, temp[1].wwww, -temp[0].xxxx;
>  7: EX2 temp[0].x, temp[1].wwww;
>  8: MOV output[0].xyz, const[0];
>  9: ADD output[0].w, -temp[0].xxxx, const[1].yyyy;
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: DP3 temp[0].x, input[0].xyz_, input[0].xyz_;
>  1: RSQ temp[0].x, |temp[0].x___|;
>  2: RCP temp[0].x, temp[0].x___;
>  3: MUL temp[0].x, temp[0].x___, const[0].w___;
>  4: MUL temp[0].x, temp[0].x___, temp[0].x___;
>  5: LG2 temp[1].w, const[1].x___;
>  6: MUL temp[1].w, temp[1].___w, -temp[0].___x;
>  7: EX2 temp[0].x, temp[1].w___;
>  8: MOV output[0].xyz, const[0].xyz_;
>  9: ADD output[0].w, -temp[0].___x, const[1].___y;
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: DP3 temp[0].x, input[0].xyz_, input[0].xyz_;
>  1: RSQ temp[0].x, |temp[0].x___|;
>  2: RCP temp[0].x, temp[0].x___;
>  3: MUL temp[0].x, temp[0].x___, const[0].w___;
>  4: MUL temp[0].x, temp[0].x___, temp[0].x___;
>  5: LG2 temp[1].w, const[1].x___;
>  6: MUL temp[1].w, temp[1].___w, -temp[0].___x;
>  7: EX2 temp[0].x, temp[1].w___;
>  8: MOV output[0].xyz, const[0].xyz_;
>  9: ADD output[0].w, -temp[0].___x, none.___1;
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: DP3 temp[0].x, input[0].xyz_, input[0].xyz_;
>  1: RSQ temp[0].x, |temp[0].x___|;
>  2: RCP temp[0].x, temp[0].x___;
>  3: MUL temp[0].x, temp[0].x___, const[0].w___;
>  4: MUL temp[0].x, temp[0].x___, temp[0].x___;
>  5: LG2 temp[1].w, const[1].x___;
>  6: MUL temp[1].w, temp[1].___w, -temp[0].___x;
>  7: EX2 temp[0].x, temp[1].w___;
>  8: MOV output[0].xyz, const[0].xyz_;
>  9: ADD output[0].w, -temp[0].___x, none.___1;
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: src0.xyz = input[0]
>     DP3 temp[0].x, src0.xyz, src0.xyz
>  1: src0.xyz = temp[0]
>     REPL_ALPHA temp[0].x
>     RSQ, |src0.x|
>  2: src0.xyz = temp[0]
>     REPL_ALPHA temp[0].x
>     RCP, src0.x
>  3: src0.xyz = temp[0], src0.w = const[0]
>     MAD temp[0].x, src0.x__, src0.w__, src0.000
>  4: src0.xyz = temp[0]
>     MAD temp[0].x, src0.x__, src0.x__, src0.000
>  5: src0.xyz = const[1]
>     LG2 temp[1].w, src0.x
>  6: src0.xyz = temp[0], src0.w = temp[1]
>     MAD temp[1].w, src0.w, -src0.x, src0.0
>  7: src0.w = temp[1]
>     REPL_ALPHA temp[0].x
>     EX2, src0.w
>  8: src0.xyz = const[0]
>     MAD color[0].xyz, src0.xyz, src0.111, src0.000
>  9: src0.xyz = temp[0]
>     MAD color[0].w, -src0.x, src0.1, src0.1
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: src0.xyz = const[0], src1.xyz = const[1]
>     MAD color[0].xyz, src0.xyz, src0.111, src0.000
>     LG2 temp[1].w, src1.x
>  1: src0.xyz = input[0]
>     DP3 temp[0].x, src0.xyz, src0.xyz
>  2: src0.xyz = temp[0]
>     REPL_ALPHA temp[0].x
>     RSQ, |src0.x|
>  3: src0.xyz = temp[0]
>     REPL_ALPHA temp[0].x
>     RCP, src0.x
>  4: src0.xyz = temp[0], src0.w = const[0]
>     MAD temp[0].x, src0.x__, src0.w__, src0.000
>  5: src0.xyz = temp[0]
>     MAD temp[0].x, src0.x__, src0.x__, src0.000
>  6: src0.xyz = temp[0], src0.w = temp[1]
>     MAD temp[1].w, src0.w, -src0.x, src0.0
>  7: src0.w = temp[1]
>     REPL_ALPHA temp[0].x
>     EX2, src0.w
>  8: src0.xyz = temp[0]
>     MAD color[0].w, -src0.x, src0.1, src0.1
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: src0.xyz = const[0], src1.xyz = const[1]
>     MAD color[0].xyz, src0.xyz, src0.111, src0.000
>     LG2 temp[1].w, src1.x
>  1: src0.xyz = temp[0]
>     DP3 temp[0].x, src0.xyz, src0.xyz
>  2: src0.xyz = temp[0]
>     REPL_ALPHA temp[0].x
>     RSQ, |src0.x|
>  3: src0.xyz = temp[0]
>     REPL_ALPHA temp[0].x
>     RCP, src0.x
>  4: src0.xyz = temp[0], src0.w = const[0]
>     MAD temp[0].x, src0.x__, src0.w__, src0.000
>  5: src0.xyz = temp[0]
>     MAD temp[0].x, src0.x__, src0.x__, src0.000
>  6: src0.xyz = temp[0], src0.w = temp[1]
>     MAD temp[1].w, src0.w, -src0.x, src0.0
>  7: src0.w = temp[1]
>     REPL_ALPHA temp[0].x
>     EX2, src0.w
>  8: src0.xyz = temp[0]
>     MAD color[0].w, -src0.x, src0.1, src0.1
>pc=47*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 8, tex_end: 0  (code_addr: 00400200)
>  0: xyz:  c0  c1  t0 -> o0.xyz               (1c000860)
>       w:  t0  t0  t0 -> t1.w                 (00840000)
>     xyz:   c0.xyz      1.0      0.0    op: 00050a80
>       w:     c1.x     c0.x     c0.x    op: 04800003
>  1: xyz:  t0  t0  t0 -> t0.x                 (00800000)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xyz   t0.xyz   t0.xxx    op: 00804000
>       w:     t0.x     t0.x     t0.x    op: 00000000
>  2: xyz:  t0  t0  t0 -> t0.x                 (00800000)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xxx   t0.xxx   t0.xxx    op: 05004081
>       w:   |t0.x|     t0.x     t0.x    op: 05800040
>  3: xyz:  t0  t0  t0 -> t0.x                 (00800000)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xxx   t0.xxx   t0.xxx    op: 05004081
>       w:     t0.x     t0.x     t0.x    op: 05000000
>  4: xyz:  t0  t0  t0 -> t0.x                 (00800000)
>       w:  c0  t0  t0 ->                      (00000020)
>     xyz:   t0.xyz   c0.www      0.0    op: 00050600
>       w:     t0.x     t0.x     t0.x    op: 00000000
>  5: xyz:  t0  t0  t0 -> t0.x                 (00800000)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xyz   t0.xyz      0.0    op: 00050000
>       w:     t0.x     t0.x     t0.x    op: 00000000
>  6: xyz:  t0  t0  t0 ->                      (00000000)
>       w:  t1  t0  t0 -> t1.w                 (00840001)
>     xyz:   t0.xxx   t0.xxx   t0.xxx    op: 00004081
>       w:     t1.w    -t0.x      0.0    op: 00041009
>  7: xyz:  t0  t0  t0 -> t0.x                 (00800000)
>       w:  t1  t0  t0 ->                      (00000001)
>     xyz:   t0.xxx   t0.xxx   t0.xxx    op: 05004081
>       w:     t1.w     t0.x     t0.x    op: 04000009
>  8: xyz:  t0  t0  t0 ->                      (00000000)
>       w:  t0  t0  t0 -> o0.w                 (01000000)
>     xyz:   t0.xxx   t0.xxx   t0.xxx    op: 00004081
>       w:    -t0.x      1.0      1.0    op: 000448a0
>r300: Initial fragment program
>FRAG
>DCL IN[0], COLOR, LINEAR
>DCL IN[1], GENERIC[0], PERSPECTIVE
>DCL IN[2], GENERIC[1], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL SAMP[1]
>DCL TEMP[0..1]
>  0: TEX TEMP[0], IN[2], SAMP[1], 2D
>  1: TEX TEMP[1], IN[1], SAMP[0], 2D
>  2: MUL TEMP[1], TEMP[1], TEMP[0]
>  3: MUL OUT[0], TEMP[1], IN[0]
>  4: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: TEX temp[0], input[2], 2D[1];
>  1: TEX temp[1], input[1], 2D[0];
>  2: MUL temp[1], temp[1], temp[0];
>  3: MUL output[0], temp[1], input[0];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: TEX temp[0], input[2], 2D[1];
>  1: TEX temp[1], input[1], 2D[0];
>  2: MUL temp[1], temp[1], temp[0];
>  3: MUL output[0], temp[1], input[0];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: TEX temp[0], input[2], 2D[1];
>  1: TEX temp[1], input[1], 2D[0];
>  2: MUL temp[1], temp[1], temp[0];
>  3: MUL output[0], temp[1], input[0];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: TEX temp[0], input[2], 2D[1];
>  1: TEX temp[1], input[1], 2D[0];
>  2: MUL temp[1], temp[1], temp[0];
>  3: MUL output[0], temp[1], input[0];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: TEX temp[0], input[2], 2D[1];
>  1: TEX temp[1], input[1], 2D[0];
>  2: MUL temp[1], temp[1], temp[0];
>  3: MUL output[0], temp[1], input[0];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: TEX temp[0], input[2].xy__, 2D[1];
>  1: TEX temp[1], input[1].xy__, 2D[0];
>  2: MUL temp[1], temp[1], temp[0];
>  3: MUL output[0], temp[1], input[0];
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: TEX temp[0], input[2].xy__, 2D[1];
>  1: TEX temp[1], input[1].xy__, 2D[0];
>  2: MUL temp[1], temp[1], temp[0];
>  3: MUL output[0], temp[1], input[0];
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: TEX temp[0], input[2].xy__, 2D[1];
>  1: TEX temp[1], input[1].xy__, 2D[0];
>  2: MUL temp[1], temp[1], temp[0];
>  3: MUL output[0], temp[1], input[0];
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: TEX temp[0], input[2].xy__, 2D[1];
>  1: TEX temp[1], input[1].xy__, 2D[0];
>  2: src0.xyz = temp[1], src0.w = temp[1], src1.xyz = temp[0], src1.w = temp[0]
>     MAD temp[1].xyz, src0.xyz, src1.xyz, src0.000
>     MAD temp[1].w, src0.w, src1.w, src0.0
>  3: src0.xyz = temp[1], src0.w = temp[1], src1.xyz = input[0], src1.w = input[0]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[1], input[1].xy__, 2D[0];
>  2: TEX temp[0], input[2].xy__, 2D[1];
>  3: src0.xyz = temp[1], src0.w = temp[1], src1.xyz = temp[0], src1.w = temp[0]
>     MAD temp[1].xyz, src0.xyz, src1.xyz, src0.000
>     MAD temp[1].w, src0.w, src1.w, src0.0
>  4: src0.xyz = temp[1], src0.w = temp[1], src1.xyz = input[0], src1.w = input[0]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[3], temp[1].xy__, 2D[0];
>  2: TEX temp[1], temp[2].xy__, 2D[1];
>  3: src0.xyz = temp[3], src0.w = temp[3], src1.xyz = temp[1], src1.w = temp[1]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src0.000
>     MAD temp[3].w, src0.w, src1.w, src0.0
>  4: src0.xyz = temp[3], src0.w = temp[3], src1.xyz = temp[0], src1.w = temp[0]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>pc=48*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 1, tex_end: 1  (code_addr: 00420040)
>  TEX:
>    TEX t3, t1, texture[0]   (000080c1)
>    TEX t1, t2, texture[1]   (00008842)
>  0: xyz:  t3  t1  t0 -> t3.xyz               (038c0043)
>       w:  t3  t1  t0 -> t3.w                 (008c0043)
>     xyz:   t3.xyz   t1.xyz      0.0    op: 00050200
>       w:     t3.w     t1.w      0.0    op: 00040509
>  1: xyz:  t3  t0  t0 -> o0.xyz               (1c000003)
>       w:  t3  t0  t0 -> o0.w                 (01000003)
>     xyz:   t3.xyz   t0.xyz      0.0    op: 00050200
>       w:     t3.w     t0.w      0.0    op: 00040509
>r300: Initial fragment program
>FRAG
>DCL IN[0], COLOR, LINEAR
>DCL IN[1], COLOR[1], LINEAR
>DCL IN[2], GENERIC[0], PERSPECTIVE
>DCL IN[3], GENERIC[1], PERSPECTIVE
>DCL IN[4], GENERIC[2], PERSPECTIVE
>DCL IN[5], GENERIC[3], PERSPECTIVE
>DCL IN[6], GENERIC[4], PERSPECTIVE
>DCL IN[7], GENERIC[5], PERSPECTIVE
>DCL IN[8], GENERIC[6], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL SAMP[1]
>DCL SAMP[3]
>DCL SAMP[4]
>DCL SAMP[5]
>DCL SAMP[6]
>DCL SAMP[7]
>DCL SAMP[8]
>DCL SAMP[9]
>DCL CONST[0]
>DCL CONST[5..8]
>DCL TEMP[0..11]
>IMM FLT32 {    2.0000,    -1.0000,     1.0000,     4.0000}
>IMM FLT32 {    0.3000,     0.5900,     0.1100,     0.0000}
>IMM FLT32 {    1.0000,     4.0000,     3.0000,     0.5000}
>  0: MAD TEMP[0].xy, IN[3], CONST[0], CONST[0].zwzw
>  1: TEX TEMP[1], IN[6], SAMP[0], 2D
>  2: TEX TEMP[2], IN[6], SAMP[1], 2D
>  3: MUL_SAT TEMP[2], TEMP[2], IN[3].wwww
>  4: DP3 TEMP[3].x, IN[8], IN[8]
>  5: RSQ TEMP[3].x, TEMP[3].xxxx
>  6: MUL TEMP[3], TEMP[3].xxxx, IN[8]
>  7: DP3 TEMP[4].x, IN[2], IN[2]
>  8: RSQ TEMP[4].x, TEMP[4].xxxx
>  9: MUL TEMP[4], TEMP[4].xxxx, IN[2]
> 10: TEX TEMP[5], IN[3], SAMP[3], 2D
> 11: MOV TEMP[5].x, TEMP[5].wwww
> 12: MAD TEMP[5], TEMP[5], IMM[0].xxxx, IMM[0].yyyy
> 13: TEX TEMP[6], TEMP[0], SAMP[4], 2D
> 14: MOV TEMP[6].x, TEMP[6].wwww
> 15: MAD TEMP[6].xy, TEMP[6], IMM[0].xxxx, IMM[0].yyyy
> 16: MAD TEMP[5].xy, TEMP[2].xxxx, TEMP[6], TEMP[5]
> 17: MOV TEMP[5].z, IMM[1].wwww
> 18: DP3 TEMP[7].x, TEMP[5], TEMP[5]
> 19: ADD TEMP[7].x, IMM[2].xxxx, -TEMP[7].xxxx
> 20: RSQ TEMP[7].x, TEMP[7].xxxx
> 21: RCP TEMP[5].z, TEMP[7].xxxx
> 22: DP3_SAT TEMP[4].x, TEMP[4], TEMP[5]
> 23: TXP TEMP[7], IN[5], SAMP[5], 2D
> 24: MUL TEMP[4], TEMP[4].xxxx, TEMP[7]
> 25: TEX TEMP[7].x, IN[4], SAMP[6], 2D
> 26: MUL TEMP[4], TEMP[4], TEMP[7].xxxx
> 27: MUL TEMP[1], TEMP[1], IN[0]
> 28: MUL TEMP[8], TEMP[1], CONST[5]
> 29: DP3 TEMP[3].x, TEMP[3], TEMP[5]
> 30: MAD_SAT TEMP[7].x, TEMP[3].xxxx, IMM[2].yyyy, -IMM[2].zzzz
> 31: MUL TEMP[7].x, TEMP[7].xxxx, TEMP[7].xxxx
> 32: MUL TEMP[7], TEMP[7].xxxx, CONST[6]
> 33: TEX TEMP[9], IN[7], SAMP[7], 2D
> 34: ADD TEMP[9], TEMP[9], TEMP[9]
> 35: TEX TEMP[10], TEMP[0], SAMP[8], 2D
> 36: ADD TEMP[10], TEMP[10], -IMM[2].wwww
> 37: MAD TEMP[9], TEMP[2].xxxx, TEMP[10], TEMP[9]
> 38: MAD TEMP[8], TEMP[7], TEMP[9], TEMP[8]
> 39: MUL TEMP[11].xyz, TEMP[4], TEMP[8]
> 40: MAD TEMP[11].xyz, TEMP[1], CONST[7], TEMP[11]
> 41: TEX TEMP[7], IN[6], SAMP[9], 2D
> 42: ADD TEMP[11], TEMP[7], TEMP[11]
> 43: LRP OUT[0].xyz, IN[1].xxxx, CONST[8], TEMP[11]
> 44: MUL TEMP[7], TEMP[7], TEMP[9]
> 45: MUL TEMP[7], TEMP[7], TEMP[4]
> 46: DP3 OUT[0].w, TEMP[7], IMM[1]
> 47: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: MAD temp[0].xy, input[3], const[0], const[0].zwzw;
>  1: TEX temp[1], input[6], 2D[0];
>  2: TEX temp[2], input[6], 2D[1];
>  3: MUL_SAT temp[2], temp[2], input[3].wwww;
>  4: DP3 temp[3].x, input[8], input[8];
>  5: RSQ temp[3].x, temp[3].xxxx;
>  6: MUL temp[3], temp[3].xxxx, input[8];
>  7: DP3 temp[4].x, input[2], input[2];
>  8: RSQ temp[4].x, temp[4].xxxx;
>  9: MUL temp[4], temp[4].xxxx, input[2];
> 10: TEX temp[5], input[3], 2D[3];
> 11: MOV temp[5].x, temp[5].wwww;
> 12: MAD temp[5], temp[5], const[9].xxxx, const[9].yyyy;
> 13: TEX temp[6], temp[0], 2D[4];
> 14: MOV temp[6].x, temp[6].wwww;
> 15: MAD temp[6].xy, temp[6], const[9].xxxx, const[9].yyyy;
> 16: MAD temp[5].xy, temp[2].xxxx, temp[6], temp[5];
> 17: MOV temp[5].z, const[10].wwww;
> 18: DP3 temp[7].x, temp[5], temp[5];
> 19: ADD temp[7].x, const[11].xxxx, -temp[7].xxxx;
> 20: RSQ temp[7].x, temp[7].xxxx;
> 21: RCP temp[5].z, temp[7].xxxx;
> 22: DP3_SAT temp[4].x, temp[4], temp[5];
> 23: TXP temp[7], input[5], 2D[5];
> 24: MUL temp[4], temp[4].xxxx, temp[7];
> 25: TEX temp[7].x, input[4], 2D[6];
> 26: MUL temp[4], temp[4], temp[7].xxxx;
> 27: MUL temp[1], temp[1], input[0];
> 28: MUL temp[8], temp[1], const[5];
> 29: DP3 temp[3].x, temp[3], temp[5];
> 30: MAD_SAT temp[7].x, temp[3].xxxx, const[11].yyyy, -const[11].zzzz;
> 31: MUL temp[7].x, temp[7].xxxx, temp[7].xxxx;
> 32: MUL temp[7], temp[7].xxxx, const[6];
> 33: TEX temp[9], input[7], 2D[7];
> 34: ADD temp[9], temp[9], temp[9];
> 35: TEX temp[10], temp[0], 2D[8];
> 36: ADD temp[10], temp[10], -const[11].wwww;
> 37: MAD temp[9], temp[2].xxxx, temp[10], temp[9];
> 38: MAD temp[8], temp[7], temp[9], temp[8];
> 39: MUL temp[11].xyz, temp[4], temp[8];
> 40: MAD temp[11].xyz, temp[1], const[7], temp[11];
> 41: TEX temp[7], input[6], 2D[9];
> 42: ADD temp[11], temp[7], temp[11];
> 43: LRP output[0].xyz, input[1].xxxx, const[8], temp[11];
> 44: MUL temp[7], temp[7], temp[9];
> 45: MUL temp[7], temp[7], temp[4];
> 46: DP3 output[0].w, temp[7], const[10];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: MAD temp[0].xy, input[3], const[0], const[0].zwzw;
>  1: TEX temp[1], input[6], 2D[0];
>  2: TEX temp[2], input[6], 2D[1];
>  3: MUL_SAT temp[2], temp[2], input[3].wwww;
>  4: DP3 temp[3].x, input[8], input[8];
>  5: RSQ temp[3].x, temp[3].xxxx;
>  6: MUL temp[3], temp[3].xxxx, input[8];
>  7: DP3 temp[4].x, input[2], input[2];
>  8: RSQ temp[4].x, temp[4].xxxx;
>  9: MUL temp[4], temp[4].xxxx, input[2];
> 10: TEX temp[5], input[3], 2D[3];
> 11: MOV temp[5].x, temp[5].wwww;
> 12: MAD temp[5], temp[5], const[9].xxxx, const[9].yyyy;
> 13: TEX temp[6], temp[0], 2D[4];
> 14: MOV temp[6].x, temp[6].wwww;
> 15: MAD temp[6].xy, temp[6], const[9].xxxx, const[9].yyyy;
> 16: MAD temp[5].xy, temp[2].xxxx, temp[6], temp[5];
> 17: MOV temp[5].z, const[10].wwww;
> 18: DP3 temp[7].x, temp[5], temp[5];
> 19: ADD temp[7].x, const[11].xxxx, -temp[7].xxxx;
> 20: RSQ temp[7].x, temp[7].xxxx;
> 21: RCP temp[5].z, temp[7].xxxx;
> 22: DP3_SAT temp[4].x, temp[4], temp[5];
> 23: TXP temp[7], input[5], 2D[5];
> 24: MUL temp[4], temp[4].xxxx, temp[7];
> 25: TEX temp[7].x, input[4], 2D[6];
> 26: MUL temp[4], temp[4], temp[7].xxxx;
> 27: MUL temp[1], temp[1], input[0];
> 28: MUL temp[8], temp[1], const[5];
> 29: DP3 temp[3].x, temp[3], temp[5];
> 30: MAD_SAT temp[7].x, temp[3].xxxx, const[11].yyyy, -const[11].zzzz;
> 31: MUL temp[7].x, temp[7].xxxx, temp[7].xxxx;
> 32: MUL temp[7], temp[7].xxxx, const[6];
> 33: TEX temp[9], input[7], 2D[7];
> 34: ADD temp[9], temp[9], temp[9];
> 35: TEX temp[10], temp[0], 2D[8];
> 36: ADD temp[10], temp[10], -const[11].wwww;
> 37: MAD temp[9], temp[2].xxxx, temp[10], temp[9];
> 38: MAD temp[8], temp[7], temp[9], temp[8];
> 39: MUL temp[11].xyz, temp[4], temp[8];
> 40: MAD temp[11].xyz, temp[1], const[7], temp[11];
> 41: TEX temp[7], input[6], 2D[9];
> 42: ADD temp[11], temp[7], temp[11];
> 43: LRP output[0].xyz, input[1].xxxx, const[8], temp[11];
> 44: MUL temp[7], temp[7], temp[9];
> 45: MUL temp[7], temp[7], temp[4];
> 46: DP3 output[0].w, temp[7], const[10];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: MAD temp[0].xy, input[3], const[0], const[0].zwzw;
>  1: TEX temp[1], input[6], 2D[0];
>  2: TEX temp[2], input[6], 2D[1];
>  3: MUL_SAT temp[2], temp[2], input[3].wwww;
>  4: DP3 temp[3].x, input[8], input[8];
>  5: RSQ temp[3].x, temp[3].xxxx;
>  6: MUL temp[3], temp[3].xxxx, input[8];
>  7: DP3 temp[4].x, input[2], input[2];
>  8: RSQ temp[4].x, temp[4].xxxx;
>  9: MUL temp[4], temp[4].xxxx, input[2];
> 10: TEX temp[5], input[3], 2D[3];
> 11: MOV temp[5].x, temp[5].wwww;
> 12: MAD temp[5], temp[5], const[9].xxxx, const[9].yyyy;
> 13: TEX temp[6], temp[0], 2D[4];
> 14: MOV temp[6].x, temp[6].wwww;
> 15: MAD temp[6].xy, temp[6], const[9].xxxx, const[9].yyyy;
> 16: MAD temp[5].xy, temp[2].xxxx, temp[6], temp[5];
> 17: MOV temp[5].z, const[10].wwww;
> 18: DP3 temp[7].x, temp[5], temp[5];
> 19: ADD temp[7].x, const[11].xxxx, -temp[7].xxxx;
> 20: RSQ temp[7].x, temp[7].xxxx;
> 21: RCP temp[5].z, temp[7].xxxx;
> 22: DP3_SAT temp[4].x, temp[4], temp[5];
> 23: TXP temp[7], input[5], 2D[5];
> 24: MUL temp[4], temp[4].xxxx, temp[7];
> 25: TEX temp[7].x, input[4], 2D[6];
> 26: MUL temp[4], temp[4], temp[7].xxxx;
> 27: MUL temp[1], temp[1], input[0];
> 28: MUL temp[8], temp[1], const[5];
> 29: DP3 temp[3].x, temp[3], temp[5];
> 30: MAD_SAT temp[7].x, temp[3].xxxx, const[11].yyyy, -const[11].zzzz;
> 31: MUL temp[7].x, temp[7].xxxx, temp[7].xxxx;
> 32: MUL temp[7], temp[7].xxxx, const[6];
> 33: TEX temp[9], input[7], 2D[7];
> 34: ADD temp[9], temp[9], temp[9];
> 35: TEX temp[10], temp[0], 2D[8];
> 36: ADD temp[10], temp[10], -const[11].wwww;
> 37: MAD temp[9], temp[2].xxxx, temp[10], temp[9];
> 38: MAD temp[8], temp[7], temp[9], temp[8];
> 39: MUL temp[11].xyz, temp[4], temp[8];
> 40: MAD temp[11].xyz, temp[1], const[7], temp[11];
> 41: TEX temp[7], input[6], 2D[9];
> 42: ADD temp[11], temp[7], temp[11];
> 43: LRP output[0].xyz, input[1].xxxx, const[8], temp[11];
> 44: MUL temp[7], temp[7], temp[9];
> 45: MUL temp[7], temp[7], temp[4];
> 46: DP3 output[0].w, temp[7], const[10];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: MAD temp[0].xy, input[3], const[0], const[0].zwzw;
>  1: TEX temp[1], input[6], 2D[0];
>  2: TEX temp[2], input[6], 2D[1];
>  3: MUL_SAT temp[2], temp[2], input[3].wwww;
>  4: DP3 temp[3].x, input[8], input[8];
>  5: RSQ temp[3].x, |temp[3].xxxx|;
>  6: MUL temp[3], temp[3].xxxx, input[8];
>  7: DP3 temp[4].x, input[2], input[2];
>  8: RSQ temp[4].x, |temp[4].xxxx|;
>  9: MUL temp[4], temp[4].xxxx, input[2];
> 10: TEX temp[5], input[3], 2D[3];
> 11: MOV temp[5].x, temp[5].wwww;
> 12: MAD temp[5], temp[5], const[9].xxxx, const[9].yyyy;
> 13: TEX temp[6], temp[0], 2D[4];
> 14: MOV temp[6].x, temp[6].wwww;
> 15: MAD temp[6].xy, temp[6], const[9].xxxx, const[9].yyyy;
> 16: MAD temp[5].xy, temp[2].xxxx, temp[6], temp[5];
> 17: MOV temp[5].z, const[10].wwww;
> 18: DP3 temp[7].x, temp[5], temp[5];
> 19: ADD temp[7].x, const[11].xxxx, -temp[7].xxxx;
> 20: RSQ temp[7].x, |temp[7].xxxx|;
> 21: RCP temp[5].z, temp[7].xxxx;
> 22: DP3_SAT temp[4].x, temp[4], temp[5];
> 23: TXP temp[7], input[5], 2D[5];
> 24: MUL temp[4], temp[4].xxxx, temp[7];
> 25: TEX temp[7].x, input[4], 2D[6];
> 26: MUL temp[4], temp[4], temp[7].xxxx;
> 27: MUL temp[1], temp[1], input[0];
> 28: MUL temp[8], temp[1], const[5];
> 29: DP3 temp[3].x, temp[3], temp[5];
> 30: MAD_SAT temp[7].x, temp[3].xxxx, const[11].yyyy, -const[11].zzzz;
> 31: MUL temp[7].x, temp[7].xxxx, temp[7].xxxx;
> 32: MUL temp[7], temp[7].xxxx, const[6];
> 33: TEX temp[9], input[7], 2D[7];
> 34: ADD temp[9], temp[9], temp[9];
> 35: TEX temp[10], temp[0], 2D[8];
> 36: ADD temp[10], temp[10], -const[11].wwww;
> 37: MAD temp[9], temp[2].xxxx, temp[10], temp[9];
> 38: MAD temp[8], temp[7], temp[9], temp[8];
> 39: MUL temp[11].xyz, temp[4], temp[8];
> 40: MAD temp[11].xyz, temp[1], const[7], temp[11];
> 41: TEX temp[7], input[6], 2D[9];
> 42: ADD temp[11], temp[7], temp[11];
> 43: ADD temp[12], const[8], -temp[11];
> 44: MAD output[0].xyz, input[1].xxxx, temp[12], temp[11];
> 45: MUL temp[7], temp[7], temp[9];
> 46: MUL temp[7], temp[7], temp[4];
> 47: DP3 output[0].w, temp[7], const[10];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: MAD temp[0].xy, input[3], const[0], const[0].zwzw;
>  1: TEX temp[1], input[6], 2D[0];
>  2: TEX temp[2], input[6], 2D[1];
>  3: MUL_SAT temp[2], temp[2], input[3].wwww;
>  4: DP3 temp[3].x, input[8], input[8];
>  5: RSQ temp[3].x, |temp[3].xxxx|;
>  6: MUL temp[3], temp[3].xxxx, input[8];
>  7: DP3 temp[4].x, input[2], input[2];
>  8: RSQ temp[4].x, |temp[4].xxxx|;
>  9: MUL temp[4], temp[4].xxxx, input[2];
> 10: TEX temp[5], input[3], 2D[3];
> 11: MOV temp[5].x, temp[5].wwww;
> 12: MAD temp[5], temp[5], const[9].xxxx, const[9].yyyy;
> 13: TEX temp[6], temp[0], 2D[4];
> 14: MOV temp[6].x, temp[6].wwww;
> 15: MAD temp[6].xy, temp[6], const[9].xxxx, const[9].yyyy;
> 16: MAD temp[5].xy, temp[2].xxxx, temp[6], temp[5];
> 17: MOV temp[5].z, const[10].wwww;
> 18: DP3 temp[7].x, temp[5], temp[5];
> 19: ADD temp[7].x, const[11].xxxx, -temp[7].xxxx;
> 20: RSQ temp[7].x, |temp[7].xxxx|;
> 21: RCP temp[5].z, temp[7].xxxx;
> 22: DP3_SAT temp[4].x, temp[4], temp[5];
> 23: TXP temp[7], input[5], 2D[5];
> 24: MUL temp[4], temp[4].xxxx, temp[7];
> 25: TEX temp[13], input[4], 2D[6];
> 26: MOV temp[7].x, temp[13];
> 27: MUL temp[4], temp[4], temp[7].xxxx;
> 28: MUL temp[1], temp[1], input[0];
> 29: MUL temp[8], temp[1], const[5];
> 30: DP3 temp[3].x, temp[3], temp[5];
> 31: MAD_SAT temp[7].x, temp[3].xxxx, const[11].yyyy, -const[11].zzzz;
> 32: MUL temp[7].x, temp[7].xxxx, temp[7].xxxx;
> 33: MUL temp[7], temp[7].xxxx, const[6];
> 34: TEX temp[9], input[7], 2D[7];
> 35: ADD temp[9], temp[9], temp[9];
> 36: TEX temp[10], temp[0], 2D[8];
> 37: ADD temp[10], temp[10], -const[11].wwww;
> 38: MAD temp[9], temp[2].xxxx, temp[10], temp[9];
> 39: MAD temp[8], temp[7], temp[9], temp[8];
> 40: MUL temp[11].xyz, temp[4], temp[8];
> 41: MAD temp[11].xyz, temp[1], const[7], temp[11];
> 42: TEX temp[7], input[6], 2D[9];
> 43: ADD temp[11], temp[7], temp[11];
> 44: ADD temp[12], const[8], -temp[11];
> 45: MAD output[0].xyz, input[1].xxxx, temp[12], temp[11];
> 46: MUL temp[7], temp[7], temp[9];
> 47: MUL temp[7], temp[7], temp[4];
> 48: DP3 output[0].w, temp[7], const[10];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: MAD temp[0].xy, input[3].xy__, const[0].xy__, const[0].zw__;
>  1: TEX temp[1].xyz, input[6].xy__, 2D[0];
>  2: TEX temp[2].x, input[6].xy__, 2D[1];
>  3: MUL_SAT temp[2].x, temp[2].x___, input[3].w___;
>  4: DP3 temp[3].x, input[8].xyz_, input[8].xyz_;
>  5: RSQ temp[3].x, |temp[3].x___|;
>  6: MUL temp[3].xyz, temp[3].xxx_, input[8].xyz_;
>  7: DP3 temp[4].x, input[2].xyz_, input[2].xyz_;
>  8: RSQ temp[4].x, |temp[4].x___|;
>  9: MUL temp[4].xyz, temp[4].xxx_, input[2].xyz_;
> 10: TEX temp[5].yw, input[3].xy__, 2D[3];
> 11: MOV temp[5].x, temp[5].w___;
> 12: MAD temp[5].xy, temp[5].xy__, const[9].xx__, const[9].yy__;
> 13: TEX temp[6].yw, temp[0].xy__, 2D[4];
> 14: MOV temp[6].x, temp[6].w___;
> 15: MAD temp[6].xy, temp[6].xy__, const[9].xx__, const[9].yy__;
> 16: MAD temp[5].xy, temp[2].xx__, temp[6].xy__, temp[5].xy__;
> 17: MOV temp[5].z, const[10].__w_;
> 18: DP3 temp[7].x, temp[5].xyz_, temp[5].xyz_;
> 19: ADD temp[7].x, const[11].x___, -temp[7].x___;
> 20: RSQ temp[7].x, |temp[7].x___|;
> 21: RCP temp[5].z, temp[7].x___;
> 22: DP3_SAT temp[4].x, temp[4].xyz_, temp[5].xyz_;
> 23: TXP temp[7].xyz, input[5].xy_w, 2D[5];
> 24: MUL temp[4].xyz, temp[4].xxx_, temp[7].xyz_;
> 25: TEX temp[13].x, input[4].xy__, 2D[6];
> 26: MOV temp[7].x, temp[13].x___;
> 27: MUL temp[4].xyz, temp[4].xyz_, temp[7].xxx_;
> 28: MUL temp[1].xyz, temp[1].xyz_, input[0].xyz_;
> 29: MUL temp[8].xyz, temp[1].xyz_, const[5].xyz_;
> 30: DP3 temp[3].x, temp[3].xyz_, temp[5].xyz_;
> 31: MAD_SAT temp[7].x, temp[3].x___, const[11].y___, -const[11].z___;
> 32: MUL temp[7].x, temp[7].x___, temp[7].x___;
> 33: MUL temp[7].xyz, temp[7].xxx_, const[6].xyz_;
> 34: TEX temp[9].xyz, input[7].xy__, 2D[7];
> 35: ADD temp[9].xyz, temp[9].xyz_, temp[9].xyz_;
> 36: TEX temp[10].xyz, temp[0].xy__, 2D[8];
> 37: ADD temp[10].xyz, temp[10].xyz_, -const[11].www_;
> 38: MAD temp[9].xyz, temp[2].xxx_, temp[10].xyz_, temp[9].xyz_;
> 39: MAD temp[8].xyz, temp[7].xyz_, temp[9].xyz_, temp[8].xyz_;
> 40: MUL temp[11].xyz, temp[4].xyz_, temp[8].xyz_;
> 41: MAD temp[11].xyz, temp[1].xyz_, const[7].xyz_, temp[11].xyz_;
> 42: TEX temp[7].xyz, input[6].xy__, 2D[9];
> 43: ADD temp[11].xyz, temp[7].xyz_, temp[11].xyz_;
> 44: ADD temp[12].xyz, const[8].xyz_, -temp[11].xyz_;
> 45: MAD output[0].xyz, input[1].xxx_, temp[12].xyz_, temp[11].xyz_;
> 46: MUL temp[7].xyz, temp[7].xyz_, temp[9].xyz_;
> 47: MUL temp[7].xyz, temp[7].xyz_, temp[4].xyz_;
> 48: DP3 output[0].w, temp[7].xyz_, const[10].xyz_;
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: MAD temp[0].xy, input[3].xy__, const[0].xy__, const[0].zw__;
>  1: TEX temp[1].xyz, input[6].xy__, 2D[0];
>  2: TEX temp[2].x, input[6].xy__, 2D[1];
>  3: MUL_SAT temp[2].x, temp[2].x___, input[3].w___;
>  4: DP3 temp[3].x, input[8].xyz_, input[8].xyz_;
>  5: RSQ temp[3].x, |temp[3].x___|;
>  6: MUL temp[3].xyz, temp[3].xxx_, input[8].xyz_;
>  7: DP3 temp[4].x, input[2].xyz_, input[2].xyz_;
>  8: RSQ temp[4].x, |temp[4].x___|;
>  9: MUL temp[4].xyz, temp[4].xxx_, input[2].xyz_;
> 10: TEX temp[5].yw, input[3].xy__, 2D[3];
> 11: MOV temp[5].x, temp[5].w___;
> 12: MAD temp[5].xy, temp[5].xy__, const[9].xx__, none.-1-1__;
> 13: TEX temp[6].yw, temp[0].xy__, 2D[4];
> 14: MOV temp[6].x, temp[6].w___;
> 15: MAD temp[6].xy, temp[6].xy__, const[9].xx__, none.-1-1__;
> 16: MAD temp[5].xy, temp[2].xx__, temp[6].xy__, temp[5].xy__;
> 17: MOV temp[5].z, none.__0_;
> 18: DP3 temp[7].x, temp[5].xyz_, temp[5].xyz_;
> 19: ADD temp[7].x, none.1___, -temp[7].x___;
> 20: RSQ temp[7].x, |temp[7].x___|;
> 21: RCP temp[5].z, temp[7].x___;
> 22: DP3_SAT temp[4].x, temp[4].xyz_, temp[5].xyz_;
> 23: TXP temp[7].xyz, input[5].xy_w, 2D[5];
> 24: MUL temp[4].xyz, temp[4].xxx_, temp[7].xyz_;
> 25: TEX temp[13].x, input[4].xy__, 2D[6];
> 26: MUL temp[4].xyz, temp[4].xyz_, temp[13].xxx_;
> 27: MUL temp[1].xyz, temp[1].xyz_, input[0].xyz_;
> 28: MUL temp[8].xyz, temp[1].xyz_, const[5].xyz_;
> 29: DP3 temp[3].x, temp[3].xyz_, temp[5].xyz_;
> 30: MAD_SAT temp[7].x, temp[3].x___, const[11].y___, -const[11].z___;
> 31: MUL temp[7].x, temp[7].x___, temp[7].x___;
> 32: MUL temp[7].xyz, temp[7].xxx_, const[6].xyz_;
> 33: TEX temp[9].xyz, input[7].xy__, 2D[7];
> 34: ADD temp[9].xyz, temp[9].xyz_, temp[9].xyz_;
> 35: TEX temp[10].xyz, temp[0].xy__, 2D[8];
> 36: ADD temp[10].xyz, temp[10].xyz_, -none.HHH_;
> 37: MAD temp[9].xyz, temp[2].xxx_, temp[10].xyz_, temp[9].xyz_;
> 38: MAD temp[8].xyz, temp[7].xyz_, temp[9].xyz_, temp[8].xyz_;
> 39: MUL temp[11].xyz, temp[4].xyz_, temp[8].xyz_;
> 40: MAD temp[11].xyz, temp[1].xyz_, const[7].xyz_, temp[11].xyz_;
> 41: TEX temp[7].xyz, input[6].xy__, 2D[9];
> 42: ADD temp[11].xyz, temp[7].xyz_, temp[11].xyz_;
> 43: ADD temp[12].xyz, const[8].xyz_, -temp[11].xyz_;
> 44: MAD output[0].xyz, input[1].xxx_, temp[12].xyz_, temp[11].xyz_;
> 45: MUL temp[7].xyz, temp[7].xyz_, temp[9].xyz_;
> 46: MUL temp[7].xyz, temp[7].xyz_, temp[4].xyz_;
> 47: DP3 output[0].w, temp[7].xyz_, const[10].xyz_;
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: MOV temp[14].x, const[0].z___;
>  1: MOV temp[14].y, const[0]._w__;
>  2: MAD temp[0].xy, input[3].xy__, const[0].xy__, temp[14].xy__;
>  3: TEX temp[1].xyz, input[6].xy__, 2D[0];
>  4: TEX temp[2].x, input[6].xy__, 2D[1];
>  5: MUL_SAT temp[2].x, temp[2].x___, input[3].w___;
>  6: DP3 temp[3].x, input[8].xyz_, input[8].xyz_;
>  7: RSQ temp[3].x, |temp[3].x___|;
>  8: MUL temp[3].xyz, temp[3].xxx_, input[8].xyz_;
>  9: DP3 temp[4].x, input[2].xyz_, input[2].xyz_;
> 10: RSQ temp[4].x, |temp[4].x___|;
> 11: MUL temp[4].xyz, temp[4].xxx_, input[2].xyz_;
> 12: TEX temp[5].yw, input[3].xy__, 2D[3];
> 13: MOV temp[5].x, temp[5].w___;
> 14: MAD temp[5].xy, temp[5].xy__, const[9].xx__, none.-1-1__;
> 15: TEX temp[6].yw, temp[0].xy__, 2D[4];
> 16: MOV temp[6].x, temp[6].w___;
> 17: MAD temp[6].xy, temp[6].xy__, const[9].xx__, none.-1-1__;
> 18: MAD temp[5].xy, temp[2].xx__, temp[6].xy__, temp[5].xy__;
> 19: MOV temp[5].z, none.__0_;
> 20: DP3 temp[7].x, temp[5].xyz_, temp[5].xyz_;
> 21: ADD temp[7].x, none.1___, -temp[7].x___;
> 22: RSQ temp[7].x, |temp[7].x___|;
> 23: RCP temp[5].z, temp[7].x___;
> 24: DP3_SAT temp[4].x, temp[4].xyz_, temp[5].xyz_;
> 25: TXP temp[7].xyz, input[5].xy_w, 2D[5];
> 26: MUL temp[4].xyz, temp[4].xxx_, temp[7].xyz_;
> 27: TEX temp[13].x, input[4].xy__, 2D[6];
> 28: MUL temp[4].xyz, temp[4].xyz_, temp[13].xxx_;
> 29: MUL temp[1].xyz, temp[1].xyz_, input[0].xyz_;
> 30: MUL temp[8].xyz, temp[1].xyz_, const[5].xyz_;
> 31: DP3 temp[3].x, temp[3].xyz_, temp[5].xyz_;
> 32: MAD_SAT temp[7].x, temp[3].x___, const[11].y___, -const[11].z___;
> 33: MUL temp[7].x, temp[7].x___, temp[7].x___;
> 34: MUL temp[7].xyz, temp[7].xxx_, const[6].xyz_;
> 35: TEX temp[9].xyz, input[7].xy__, 2D[7];
> 36: ADD temp[9].xyz, temp[9].xyz_, temp[9].xyz_;
> 37: TEX temp[10].xyz, temp[0].xy__, 2D[8];
> 38: ADD temp[10].xyz, temp[10].xyz_, -none.HHH_;
> 39: MAD temp[9].xyz, temp[2].xxx_, temp[10].xyz_, temp[9].xyz_;
> 40: MAD temp[8].xyz, temp[7].xyz_, temp[9].xyz_, temp[8].xyz_;
> 41: MUL temp[11].xyz, temp[4].xyz_, temp[8].xyz_;
> 42: MAD temp[11].xyz, temp[1].xyz_, const[7].xyz_, temp[11].xyz_;
> 43: TEX temp[7].xyz, input[6].xy__, 2D[9];
> 44: ADD temp[11].xyz, temp[7].xyz_, temp[11].xyz_;
> 45: ADD temp[12].xyz, const[8].xyz_, -temp[11].xyz_;
> 46: MAD output[0].xyz, input[1].xxx_, temp[12].xyz_, temp[11].xyz_;
> 47: MUL temp[7].xyz, temp[7].xyz_, temp[9].xyz_;
> 48: MUL temp[7].xyz, temp[7].xyz_, temp[4].xyz_;
> 49: DP3 output[0].w, temp[7].xyz_, const[10].xyz_;
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: src0.xyz = const[0]
>     MAD temp[14].x, src0.z__, src0.111, src0.000
>  1: src0.w = const[0]
>     MAD temp[14].y, src0._w_, src0.111, src0.000
>  2: src0.xyz = input[3], src1.xyz = const[0], src2.xyz = temp[14]
>     MAD temp[0].xy, src0.xy_, src1.xy_, src2.xy_
>  3: TEX temp[1].xyz, input[6].xy__, 2D[0];
>  4: TEX temp[2].x, input[6].xy__, 2D[1];
>  5: src0.xyz = temp[2], src0.w = input[3]
>     MAD_SAT temp[2].x, src0.x__, src0.w__, src0.000
>  6: src0.xyz = input[8]
>     DP3 temp[3].x, src0.xyz, src0.xyz
>  7: src0.xyz = temp[3]
>     REPL_ALPHA temp[3].x
>     RSQ, |src0.x|
>  8: src0.xyz = temp[3], src1.xyz = input[8]
>     MAD temp[3].xyz, src0.xxx, src1.xyz, src0.000
>  9: src0.xyz = input[2]
>     DP3 temp[4].x, src0.xyz, src0.xyz
> 10: src0.xyz = temp[4]
>     REPL_ALPHA temp[4].x
>     RSQ, |src0.x|
> 11: src0.xyz = temp[4], src1.xyz = input[2]
>     MAD temp[4].xyz, src0.xxx, src1.xyz, src0.000
> 12: TEX temp[5].yw, input[3].xy__, 2D[3];
> 13: src0.w = temp[5]
>     MAD temp[5].x, src0.w__, src0.111, src0.000
> 14: src0.xyz = temp[5], src1.xyz = const[9]
>     MAD temp[5].xy, src0.xy_, src1.xx_, -src0.11_
> 15: TEX temp[6].yw, temp[0].xy__, 2D[4];
> 16: src0.w = temp[6]
>     MAD temp[6].x, src0.w__, src0.111, src0.000
> 17: src0.xyz = temp[6], src1.xyz = const[9]
>     MAD temp[6].xy, src0.xy_, src1.xx_, -src0.11_
> 18: src0.xyz = temp[2], src1.xyz = temp[6], src2.xyz = temp[5]
>     MAD temp[5].xy, src0.xx_, src1.xy_, src2.xy_
> 19: 
>     MAD temp[5].z, src0.__0, src0.111, src0.000
> 20: src0.xyz = temp[5]
>     DP3 temp[7].x, src0.xyz, src0.xyz
> 21: src0.xyz = temp[7]
>     MAD temp[7].x, src0.1__, src0.111, -src0.x__
> 22: src0.xyz = temp[7]
>     REPL_ALPHA temp[7].x
>     RSQ, |src0.x|
> 23: src0.xyz = temp[7]
>     REPL_ALPHA temp[5].z
>     RCP, src0.x
> 24: src0.xyz = temp[4], src1.xyz = temp[5]
>     DP3_SAT temp[4].x, src0.xyz, src1.xyz
> 25: TXP temp[7].xyz, input[5].xy_w, 2D[5];
> 26: src0.xyz = temp[4], src1.xyz = temp[7]
>     MAD temp[4].xyz, src0.xxx, src1.xyz, src0.000
> 27: TEX temp[13].x, input[4].xy__, 2D[6];
> 28: src0.xyz = temp[4], src1.xyz = temp[13]
>     MAD temp[4].xyz, src0.xyz, src1.xxx, src0.000
> 29: src0.xyz = temp[1], src1.xyz = input[0]
>     MAD temp[1].xyz, src0.xyz, src1.xyz, src0.000
> 30: src0.xyz = temp[1], src1.xyz = const[5]
>     MAD temp[8].xyz, src0.xyz, src1.xyz, src0.000
> 31: src0.xyz = temp[3], src1.xyz = temp[5]
>     DP3 temp[3].x, src0.xyz, src1.xyz
> 32: src0.xyz = temp[3], src1.xyz = const[11]
>     MAD_SAT temp[7].x, src0.x__, src1.y__, -src1.z__
> 33: src0.xyz = temp[7]
>     MAD temp[7].x, src0.x__, src0.x__, src0.000
> 34: src0.xyz = temp[7], src1.xyz = const[6]
>     MAD temp[7].xyz, src0.xxx, src1.xyz, src0.000
> 35: TEX temp[9].xyz, input[7].xy__, 2D[7];
> 36: src0.xyz = temp[9]
>     MAD temp[9].xyz, src0.xyz, src0.111, src0.xyz
> 37: TEX temp[10].xyz, temp[0].xy__, 2D[8];
> 38: src0.xyz = temp[10]
>     MAD temp[10].xyz, src0.xyz, src0.111, -src0.HHH
> 39: src0.xyz = temp[2], src1.xyz = temp[10], src2.xyz = temp[9]
>     MAD temp[9].xyz, src0.xxx, src1.xyz, src2.xyz
> 40: src0.xyz = temp[7], src1.xyz = temp[9], src2.xyz = temp[8]
>     MAD temp[8].xyz, src0.xyz, src1.xyz, src2.xyz
> 41: src0.xyz = temp[4], src1.xyz = temp[8]
>     MAD temp[11].xyz, src0.xyz, src1.xyz, src0.000
> 42: src0.xyz = temp[1], src1.xyz = const[7], src2.xyz = temp[11]
>     MAD temp[11].xyz, src0.xyz, src1.xyz, src2.xyz
> 43: TEX temp[7].xyz, input[6].xy__, 2D[9];
> 44: src0.xyz = temp[7], src1.xyz = temp[11]
>     MAD temp[11].xyz, src0.xyz, src0.111, src1.xyz
> 45: src0.xyz = const[8], src1.xyz = temp[11]
>     MAD temp[12].xyz, src0.xyz, src0.111, -src1.xyz
> 46: src0.xyz = input[1], src1.xyz = temp[12], src2.xyz = temp[11]
>     MAD color[0].xyz, src0.xxx, src1.xyz, src2.xyz
> 47: src0.xyz = temp[7], src1.xyz = temp[9]
>     MAD temp[7].xyz, src0.xyz, src1.xyz, src0.000
> 48: src0.xyz = temp[7], src1.xyz = temp[4]
>     MAD temp[7].xyz, src0.xyz, src1.xyz, src0.000
> 49: src0.xyz = temp[7], src1.xyz = const[10]
>     DP3, src0.xyz, src1.xyz
>     DP3 color[0].w, src0._, src0._
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[9].xyz, input[7].xy__, 2D[7];
>  2: TEX temp[13].x, input[4].xy__, 2D[6];
>  3: TEX temp[5].yw, input[3].xy__, 2D[3];
>  4: TEX temp[2].x, input[6].xy__, 2D[1];
>  5: TEX temp[1].xyz, input[6].xy__, 2D[0];
>  6: src0.xyz = temp[1], src1.xyz = input[0]
>     MAD temp[1].xyz, src0.xyz, src1.xyz, src0.000
>  7: src0.xyz = temp[1], src1.xyz = const[5]
>     MAD temp[8].xyz, src0.xyz, src1.xyz, src0.000
>  8: src0.xyz = temp[2], src0.w = input[3]
>     MAD_SAT temp[2].x, src0.x__, src0.w__, src0.000
>  9: src0.w = temp[5]
>     MAD temp[5].x, src0.w__, src0.111, src0.000
> 10: src0.xyz = temp[5], src1.xyz = const[9]
>     MAD temp[5].xy, src0.xy_, src1.xx_, -src0.11_
> 11: src0.xyz = temp[9]
>     MAD temp[9].xyz, src0.xyz, src0.111, src0.xyz
> 12: 
>     MAD temp[5].z, src0.__0, src0.111, src0.000
> 13: src0.xyz = input[2]
>     DP3 temp[4].x, src0.xyz, src0.xyz
> 14: src0.xyz = temp[4]
>     REPL_ALPHA temp[4].x
>     RSQ, |src0.x|
> 15: src0.xyz = temp[4], src1.xyz = input[2]
>     MAD temp[4].xyz, src0.xxx, src1.xyz, src0.000
> 16: src0.xyz = input[8]
>     DP3 temp[3].x, src0.xyz, src0.xyz
> 17: src0.xyz = temp[3]
>     REPL_ALPHA temp[3].x
>     RSQ, |src0.x|
> 18: src0.xyz = temp[3], src1.xyz = input[8]
>     MAD temp[3].xyz, src0.xxx, src1.xyz, src0.000
> 19: src0.w = const[0]
>     MAD temp[14].y, src0._w_, src0.111, src0.000
> 20: src0.xyz = const[0]
>     MAD temp[14].x, src0.z__, src0.111, src0.000
> 21: src0.xyz = input[3], src1.xyz = const[0], src2.xyz = temp[14]
>     MAD temp[0].xy, src0.xy_, src1.xy_, src2.xy_
> 22: BEGIN_TEX;
> 23: TEX temp[6].yw, temp[0].xy__, 2D[4];
> 24: TEX temp[10].xyz, temp[0].xy__, 2D[8];
> 25: src0.xyz = temp[10]
>     MAD temp[10].xyz, src0.xyz, src0.111, -src0.HHH
> 26: src0.xyz = temp[2], src1.xyz = temp[10], src2.xyz = temp[9]
>     MAD temp[9].xyz, src0.xxx, src1.xyz, src2.xyz
> 27: src0.w = temp[6]
>     MAD temp[6].x, src0.w__, src0.111, src0.000
> 28: src0.xyz = temp[6], src1.xyz = const[9]
>     MAD temp[6].xy, src0.xy_, src1.xx_, -src0.11_
> 29: src0.xyz = temp[2], src1.xyz = temp[6], src2.xyz = temp[5]
>     MAD temp[5].xy, src0.xx_, src1.xy_, src2.xy_
> 30: src0.xyz = temp[5]
>     DP3 temp[7].x, src0.xyz, src0.xyz
> 31: src0.xyz = temp[7]
>     MAD temp[7].x, src0.1__, src0.111, -src0.x__
> 32: src0.xyz = temp[7]
>     REPL_ALPHA temp[7].x
>     RSQ, |src0.x|
> 33: src0.xyz = temp[7]
>     REPL_ALPHA temp[5].z
>     RCP, src0.x
> 34: src0.xyz = temp[4], src1.xyz = temp[5]
>     DP3_SAT temp[4].x, src0.xyz, src1.xyz
> 35: src0.xyz = temp[3], src1.xyz = temp[5]
>     DP3 temp[3].x, src0.xyz, src1.xyz
> 36: BEGIN_TEX;
> 37: TXP temp[7].xyz, input[5].xy_w, 2D[5];
> 38: src0.xyz = temp[4], src1.xyz = temp[7]
>     MAD temp[4].xyz, src0.xxx, src1.xyz, src0.000
> 39: src0.xyz = temp[4], src1.xyz = temp[13]
>     MAD temp[4].xyz, src0.xyz, src1.xxx, src0.000
> 40: src0.xyz = temp[3], src1.xyz = const[11]
>     MAD_SAT temp[7].x, src0.x__, src1.y__, -src1.z__
> 41: src0.xyz = temp[7]
>     MAD temp[7].x, src0.x__, src0.x__, src0.000
> 42: src0.xyz = temp[7], src1.xyz = const[6]
>     MAD temp[7].xyz, src0.xxx, src1.xyz, src0.000
> 43: src0.xyz = temp[7], src1.xyz = temp[9], src2.xyz = temp[8]
>     MAD temp[8].xyz, src0.xyz, src1.xyz, src2.xyz
> 44: src0.xyz = temp[4], src1.xyz = temp[8]
>     MAD temp[11].xyz, src0.xyz, src1.xyz, src0.000
> 45: src0.xyz = temp[1], src1.xyz = const[7], src2.xyz = temp[11]
>     MAD temp[11].xyz, src0.xyz, src1.xyz, src2.xyz
> 46: BEGIN_TEX;
> 47: TEX temp[7].xyz, input[6].xy__, 2D[9];
> 48: src0.xyz = temp[7], src1.xyz = temp[11]
>     MAD temp[11].xyz, src0.xyz, src0.111, src1.xyz
> 49: src0.xyz = const[8], src1.xyz = temp[11]
>     MAD temp[12].xyz, src0.xyz, src0.111, -src1.xyz
> 50: src0.xyz = input[1], src1.xyz = temp[12], src2.xyz = temp[11]
>     MAD color[0].xyz, src0.xxx, src1.xyz, src2.xyz
> 51: src0.xyz = temp[7], src1.xyz = temp[9]
>     MAD temp[7].xyz, src0.xyz, src1.xyz, src0.000
> 52: src0.xyz = temp[7], src1.xyz = temp[4]
>     MAD temp[7].xyz, src0.xyz, src1.xyz, src0.000
> 53: src0.xyz = temp[7], src1.xyz = const[10]
>     DP3, src0.xyz, src1.xyz
>     DP3 color[0].w, src0._, src0._
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[12].xyz, temp[7].xy__, 2D[7];
>  2: TEX temp[13].x, temp[4].xy__, 2D[6];
>  3: TEX temp[10].yw, temp[3].xy__, 2D[3];
>  4: TEX temp[7].x, temp[6].xy__, 2D[1];
>  5: TEX temp[4].xyz, temp[6].xy__, 2D[0];
>  6: src0.xyz = temp[4], src1.xyz = temp[0]
>     MAD temp[4].xyz, src0.xyz, src1.xyz, src0.000
>  7: src0.xyz = temp[4], src1.xyz = const[5]
>     MAD temp[11].xyz, src0.xyz, src1.xyz, src0.000
>  8: src0.xyz = temp[7], src0.w = temp[3]
>     MAD_SAT temp[7].x, src0.x__, src0.w__, src0.000
>  9: src0.w = temp[10]
>     MAD temp[10].x, src0.w__, src0.111, src0.000
> 10: src0.xyz = temp[10], src1.xyz = const[9]
>     MAD temp[10].xy, src0.xy_, src1.xx_, -src0.11_
> 11: src0.xyz = temp[12]
>     MAD temp[12].xyz, src0.xyz, src0.111, src0.xyz
> 12: 
>     MAD temp[10].z, src0.__0, src0.111, src0.000
> 13: src0.xyz = temp[2]
>     DP3 temp[9].x, src0.xyz, src0.xyz
> 14: src0.xyz = temp[9]
>     REPL_ALPHA temp[9].x
>     RSQ, |src0.x|
> 15: src0.xyz = temp[9], src1.xyz = temp[2]
>     MAD temp[9].xyz, src0.xxx, src1.xyz, src0.000
> 16: src0.xyz = temp[8]
>     DP3 temp[2].x, src0.xyz, src0.xyz
> 17: src0.xyz = temp[2]
>     REPL_ALPHA temp[2].x
>     RSQ, |src0.x|
> 18: src0.xyz = temp[2], src1.xyz = temp[8]
>     MAD temp[2].xyz, src0.xxx, src1.xyz, src0.000
> 19: src0.w = const[0]
>     MAD temp[0].y, src0._w_, src0.111, src0.000
> 20: src0.xyz = const[0]
>     MAD temp[0].x, src0.z__, src0.111, src0.000
> 21: src0.xyz = temp[3], src1.xyz = const[0], src2.xyz = temp[0]
>     MAD temp[0].xy, src0.xy_, src1.xy_, src2.xy_
> 22: BEGIN_TEX;
> 23: TEX temp[3].yw, temp[0].xy__, 2D[4];
> 24: TEX temp[0].xyz, temp[0].xy__, 2D[8];
> 25: src0.xyz = temp[0]
>     MAD temp[0].xyz, src0.xyz, src0.111, -src0.HHH
> 26: src0.xyz = temp[7], src1.xyz = temp[0], src2.xyz = temp[12]
>     MAD temp[12].xyz, src0.xxx, src1.xyz, src2.xyz
> 27: src0.w = temp[3]
>     MAD temp[3].x, src0.w__, src0.111, src0.000
> 28: src0.xyz = temp[3], src1.xyz = const[9]
>     MAD temp[3].xy, src0.xy_, src1.xx_, -src0.11_
> 29: src0.xyz = temp[7], src1.xyz = temp[3], src2.xyz = temp[10]
>     MAD temp[10].xy, src0.xx_, src1.xy_, src2.xy_
> 30: src0.xyz = temp[10]
>     DP3 temp[0].x, src0.xyz, src0.xyz
> 31: src0.xyz = temp[0]
>     MAD temp[0].x, src0.1__, src0.111, -src0.x__
> 32: src0.xyz = temp[0]
>     REPL_ALPHA temp[0].x
>     RSQ, |src0.x|
> 33: src0.xyz = temp[0]
>     REPL_ALPHA temp[10].z
>     RCP, src0.x
> 34: src0.xyz = temp[9], src1.xyz = temp[10]
>     DP3_SAT temp[9].x, src0.xyz, src1.xyz
> 35: src0.xyz = temp[2], src1.xyz = temp[10]
>     DP3 temp[2].x, src0.xyz, src1.xyz
> 36: BEGIN_TEX;
> 37: TXP temp[0].xyz, temp[5].xy_w, 2D[5];
> 38: src0.xyz = temp[9], src1.xyz = temp[0]
>     MAD temp[9].xyz, src0.xxx, src1.xyz, src0.000
> 39: src0.xyz = temp[9], src1.xyz = temp[13]
>     MAD temp[9].xyz, src0.xyz, src1.xxx, src0.000
> 40: src0.xyz = temp[2], src1.xyz = const[11]
>     MAD_SAT temp[0].x, src0.x__, src1.y__, -src1.z__
> 41: src0.xyz = temp[0]
>     MAD temp[0].x, src0.x__, src0.x__, src0.000
> 42: src0.xyz = temp[0], src1.xyz = const[6]
>     MAD temp[0].xyz, src0.xxx, src1.xyz, src0.000
> 43: src0.xyz = temp[0], src1.xyz = temp[12], src2.xyz = temp[11]
>     MAD temp[11].xyz, src0.xyz, src1.xyz, src2.xyz
> 44: src0.xyz = temp[9], src1.xyz = temp[11]
>     MAD temp[2].xyz, src0.xyz, src1.xyz, src0.000
> 45: src0.xyz = temp[4], src1.xyz = const[7], src2.xyz = temp[2]
>     MAD temp[2].xyz, src0.xyz, src1.xyz, src2.xyz
> 46: BEGIN_TEX;
> 47: TEX temp[0].xyz, temp[6].xy__, 2D[9];
> 48: src0.xyz = temp[0], src1.xyz = temp[2]
>     MAD temp[2].xyz, src0.xyz, src0.111, src1.xyz
> 49: src0.xyz = const[8], src1.xyz = temp[2]
>     MAD temp[3].xyz, src0.xyz, src0.111, -src1.xyz
> 50: src0.xyz = temp[1], src1.xyz = temp[3], src2.xyz = temp[2]
>     MAD color[0].xyz, src0.xxx, src1.xyz, src2.xyz
> 51: src0.xyz = temp[0], src1.xyz = temp[12]
>     MAD temp[0].xyz, src0.xyz, src1.xyz, src0.000
> 52: src0.xyz = temp[0], src1.xyz = temp[9]
>     MAD temp[0].xyz, src0.xyz, src1.xyz, src0.000
> 53: src0.xyz = temp[0], src1.xyz = const[10]
>     DP3, src0.xyz, src1.xyz
>     DP3 color[0].w, src0._, src0._
>pc=49*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 15, tex_end: 4  (code_addr: 000803c0)
>  TEX:
>    TEX t12, t7, texture[7]   (0000bb07)
>    TEX t13, t4, texture[6]   (0000b344)
>    TEX t10, t3, texture[3]   (00009a83)
>    TEX t7, t6, texture[1]   (000089c6)
>    TEX t4, t6, texture[0]   (00008106)
>  0: xyz:  t4  t0  t0 -> t4.xyz               (03900004)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t4.xyz   t0.xyz      0.0    op: 00050200
>       w:     t4.x     t4.x     t4.x    op: 00000000
>  1: xyz:  t4  c5  t0 -> t11.xyz              (03ac0944)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t4.xyz   c5.xyz      0.0    op: 00050200
>       w:     t4.x     t4.x     t4.x    op: 00000000
>  2: xyz:  t7  t0  t0 -> t7.x                 (009c0007)
>       w:  t3  t0  t0 ->                      (00000003)
>     xyz:   t7.xyz   t3.www      0.0    op: 40050600
>       w:     t7.x     t7.x     t7.x    op: 00000000
>  3: xyz:  t0  t0  t0 -> t10.x                (00a80000)
>       w: t10  t0  t0 ->                      (0000000a)
>     xyz:  t10.www      1.0      0.0    op: 00050a8c
>       w:     t0.x     t0.x     t0.x    op: 00000000
>  4: xyz: t10  c9  t0 -> t10.xy               (01a80a4a)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:  t10.xyz   c9.xxx     -1.0    op: 000d4280
>       w:    t10.x    t10.x    t10.x    op: 00000000
>  5: xyz: t12  t0  t0 -> t12.xyz              (03b0000c)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:  t12.xyz      1.0  t12.xyz    op: 00000a80
>       w:    t12.x    t12.x    t12.x    op: 00000000
>  6: xyz:  t0  t0  t0 -> t10.z                (02280000)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:      0.0      1.0      0.0    op: 00050a94
>       w:     t0.x     t0.x     t0.x    op: 00000000
>  7: xyz:  t2  t0  t0 -> t9.x                 (00a40002)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz   t2.xyz   t2.xxx    op: 00804000
>       w:     t2.x     t2.x     t2.x    op: 00000000
>  8: xyz:  t9  t0  t0 -> t9.x                 (00a40009)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t9.xxx   t9.xxx   t9.xxx    op: 05004081
>       w:   |t9.x|     t9.x     t9.x    op: 05800040
>  9: xyz:  t9  t2  t0 -> t9.xyz               (03a40089)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t9.xxx   t2.xyz      0.0    op: 00050201
>       w:     t9.x     t9.x     t9.x    op: 00000000
> 10: xyz:  t8  t0  t0 -> t2.x                 (00880008)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t8.xyz   t8.xyz   t8.xxx    op: 00804000
>       w:     t8.x     t8.x     t8.x    op: 00000000
> 11: xyz:  t2  t0  t0 -> t2.x                 (00880002)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xxx   t2.xxx   t2.xxx    op: 05004081
>       w:   |t2.x|     t2.x     t2.x    op: 05800040
> 12: xyz:  t2  t8  t0 -> t2.xyz               (03880202)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xxx   t8.xyz      0.0    op: 00050201
>       w:     t2.x     t2.x     t2.x    op: 00000000
> 13: xyz:  t0  t0  t0 -> t0.y                 (01000000)
>       w:  c0  t0  t0 ->                      (00000020)
>     xyz:   c0.www      1.0      0.0    op: 00050a8c
>       w:     t0.x     t0.x     t0.x    op: 00000000
> 14: xyz:  c0  t0  t0 -> t0.x                 (00800020)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   c0.zzz      1.0      0.0    op: 00050a83
>       w:     c0.x     c0.x     c0.x    op: 00000000
> 15: xyz:  t3  c0  t0 -> t0.xy                (01800803)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t3.xyz   c0.xyz   t0.xyz    op: 00020200
>       w:     t3.x     t3.x     t3.x    op: 00000000
>NODE 1: alu_offset: 16, tex_offset: 5, alu_end: 10, tex_end: 1  (code_addr: 00025290)
>  TEX:
>    TEX t3, t0, texture[4]   (0000a0c0)
>    TEX t0, t0, texture[8]   (0000c000)
> 16: xyz:  t0  t0  t0 -> t0.xyz               (03800000)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xyz      1.0     -0.5    op: 000d8a80
>       w:     t0.x     t0.x     t0.x    op: 00000000
> 17: xyz:  t7  t0 t12 -> t12.xyz              (03b0c007)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t7.xxx   t0.xyz  t12.xyz    op: 00020201
>       w:     t7.x     t7.x     t7.x    op: 00000000
> 18: xyz:  t0  t0  t0 -> t3.x                 (008c0000)
>       w:  t3  t0  t0 ->                      (00000003)
>     xyz:   t3.www      1.0      0.0    op: 00050a8c
>       w:     t0.x     t0.x     t0.x    op: 00000000
> 19: xyz:  t3  c9  t0 -> t3.xy                (018c0a43)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t3.xyz   c9.xxx     -1.0    op: 000d4280
>       w:     t3.x     t3.x     t3.x    op: 00000000
> 20: xyz:  t7  t3 t10 -> t10.xy               (01a8a0c7)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t7.xxx   t3.xyz  t10.xyz    op: 00020201
>       w:     t7.x     t7.x     t7.x    op: 00000000
> 21: xyz: t10  t0  t0 -> t0.x                 (0080000a)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:  t10.xyz  t10.xyz  t10.xxx    op: 00804000
>       w:    t10.x    t10.x    t10.x    op: 00000000
> 22: xyz:  t0  t0  t0 -> t0.x                 (00800000)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:      1.0      1.0  -t0.xyz    op: 00080a95
>       w:     t0.x     t0.x     t0.x    op: 00000000
> 23: xyz:  t0  t0  t0 -> t0.x                 (00800000)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xxx   t0.xxx   t0.xxx    op: 05004081
>       w:   |t0.x|     t0.x     t0.x    op: 05800040
> 24: xyz:  t0  t0  t0 -> t10.z                (02280000)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xxx   t0.xxx   t0.xxx    op: 05004081
>       w:     t0.x     t0.x     t0.x    op: 05000000
> 25: xyz:  t9 t10  t0 -> t9.x                 (00a40289)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t9.xyz  t10.xyz   t9.xxx    op: 40804200
>       w:     t9.x     t9.x     t9.x    op: 00000000
> 26: xyz:  t2 t10  t0 -> t2.x                 (00880282)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz  t10.xyz   t2.xxx    op: 00804200
>       w:     t2.x     t2.x     t2.x    op: 00000000
>NODE 2: alu_offset: 27, tex_offset: 7, alu_end: 7, tex_end: 0  (code_addr: 000071db)
>  TEX:
>    TXP t0, t5, texture[5]   (0001a805)
> 27: xyz:  t9  t0  t0 -> t9.xyz               (03a40009)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t9.xxx   t0.xyz      0.0    op: 00050201
>       w:     t9.x     t9.x     t9.x    op: 00000000
> 28: xyz:  t9 t13  t0 -> t9.xyz               (03a40349)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t9.xyz  t13.xxx      0.0    op: 00050280
>       w:     t9.x     t9.x     t9.x    op: 00000000
> 29: xyz:  t2 c11  t0 -> t0.x                 (00800ac2)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz  c11.yyy -c11.zzz    op: 4009c300
>       w:     t2.x     t2.x     t2.x    op: 00000000
> 30: xyz:  t0  t0  t0 -> t0.x                 (00800000)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xyz   t0.xyz      0.0    op: 00050000
>       w:     t0.x     t0.x     t0.x    op: 00000000
> 31: xyz:  t0  c6  t0 -> t0.xyz               (03800980)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xxx   c6.xyz      0.0    op: 00050201
>       w:     t0.x     t0.x     t0.x    op: 00000000
> 32: xyz:  t0 t12 t11 -> t11.xyz              (03acb300)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xyz  t12.xyz  t11.xyz    op: 00020200
>       w:     t0.x     t0.x     t0.x    op: 00000000
> 33: xyz:  t9 t11  t0 -> t2.xyz               (038802c9)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t9.xyz  t11.xyz      0.0    op: 00050200
>       w:     t9.x     t9.x     t9.x    op: 00000000
> 34: xyz:  t4  c7  t2 -> t2.xyz               (038829c4)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t4.xyz   c7.xyz   t2.xyz    op: 00020200
>       w:     t4.x     t4.x     t4.x    op: 00000000
>NODE 3: alu_offset: 35, tex_offset: 8, alu_end: 5, tex_end: 0  (code_addr: 00408163)
>  TEX:
>    TEX t0, t6, texture[9]   (0000c806)
> 35: xyz:  t0  t2  t0 -> t2.xyz               (03880080)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xyz      1.0   t2.xyz    op: 00010a80
>       w:     t0.x     t0.x     t0.x    op: 00000000
> 36: xyz:  c8  t2  t0 -> t3.xyz               (038c00a8)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   c8.xyz      1.0  -t2.xyz    op: 00090a80
>       w:     c8.x     c8.x     c8.x    op: 00000000
> 37: xyz:  t1  t3  t2 -> o0.xyz               (1c0020c1)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xxx   t3.xyz   t2.xyz    op: 00020201
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 38: xyz:  t0 t12  t0 -> t0.xyz               (03800300)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xyz  t12.xyz      0.0    op: 00050200
>       w:     t0.x     t0.x     t0.x    op: 00000000
> 39: xyz:  t0  t9  t0 -> t0.xyz               (03800240)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xyz   t9.xyz      0.0    op: 00050200
>       w:     t0.x     t0.x     t0.x    op: 00000000
> 40: xyz:  t0 c10  t0 ->                      (00000a80)
>       w:  t0  t0  t0 -> o0.w                 (01000000)
>     xyz:   t0.xyz  c10.xyz   t0.xxx    op: 00804200
>       w:      1.0      1.0     t0.x    op: 00800891
>r300: Initial fragment program
>FRAG
>DCL IN[0], COLOR[1], LINEAR
>DCL IN[1], GENERIC[0], PERSPECTIVE
>DCL IN[2], GENERIC[1], PERSPECTIVE
>DCL IN[3], GENERIC[2], PERSPECTIVE
>DCL IN[4], GENERIC[3], PERSPECTIVE
>DCL IN[5], GENERIC[4], PERSPECTIVE
>DCL IN[6], GENERIC[5], PERSPECTIVE
>DCL IN[7], GENERIC[6], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[1]
>DCL SAMP[2]
>DCL SAMP[3]
>DCL SAMP[4]
>DCL SAMP[5]
>DCL CONST[4..8]
>DCL TEMP[0..9]
>IMM FLT32 {    2.0000,    -1.0000,     1.0000,     4.0000}
>IMM FLT32 {    0.3000,     0.5900,     0.1100,     0.0000}
>IMM FLT32 {    1.0000,     4.0000,     3.0000,     0.0000}
>  0: DP3 TEMP[0].x, IN[7], IN[7]
>  1: RSQ TEMP[0].x, TEMP[0].xxxx
>  2: MUL TEMP[0], TEMP[0].xxxx, IN[7]
>  3: DP3 TEMP[1].x, IN[1], IN[1]
>  4: RSQ TEMP[1].x, TEMP[1].xxxx
>  5: MUL TEMP[1], TEMP[1].xxxx, IN[1]
>  6: TEX TEMP[2], IN[2], SAMP[1], 2D
>  7: MOV TEMP[2].x, TEMP[2].wwww
>  8: MAD TEMP[2], TEMP[2], IMM[0].xxxx, IMM[0].yyyy
>  9: MOV TEMP[2].z, IMM[1].wwww
> 10: DP3 TEMP[3].x, TEMP[2], TEMP[2]
> 11: ADD TEMP[3].x, IMM[2].xxxx, -TEMP[3].xxxx
> 12: RSQ TEMP[3].x, TEMP[3].xxxx
> 13: RCP TEMP[2].z, TEMP[3].xxxx
> 14: MOV TEMP[4], TEMP[1].zzzz
> 15: TEX TEMP[5], IN[5], SAMP[2], 2D
> 16: MOV TEMP[6], TEMP[5]
> 17: DP3 TEMP[7].x, TEMP[1], -TEMP[2]
> 18: MUL TEMP[7].x, TEMP[7].xxxx, TEMP[5].wwww
> 19: MUL TEMP[7].x, TEMP[7].xxxx, CONST[4].wwww
> 20: DP3 TEMP[8].x, TEMP[1], TEMP[2]
> 21: CMP TEMP[1].x, TEMP[4].xxxx, TEMP[7].xxxx, TEMP[8].xxxx
> 22: TXP TEMP[3], IN[4], SAMP[3], 2D
> 23: MUL TEMP[1], TEMP[1].xxxx, TEMP[3]
> 24: TEX TEMP[3].x, IN[3], SAMP[4], 2D
> 25: MUL TEMP[1], TEMP[1], TEMP[3].xxxx
> 26: MUL TEMP[9], TEMP[5], CONST[5]
> 27: MUL TEMP[3].xyz, TEMP[9], CONST[4]
> 28: CMP TEMP[9].xyz, TEMP[4].xxxx, TEMP[3], TEMP[9]
> 29: DP3 TEMP[0].x, TEMP[0], TEMP[2]
> 30: MAD_SAT TEMP[3].x, TEMP[0].xxxx, IMM[2].yyyy, -IMM[2].zzzz
> 31: MUL TEMP[3].x, TEMP[3].xxxx, TEMP[3].xxxx
> 32: MUL TEMP[3], TEMP[3].xxxx, CONST[6]
> 33: CMP TEMP[3], TEMP[4].xxxx, IMM[1].wwww, TEMP[3]
> 34: TEX TEMP[5], IN[6], SAMP[5], 2D
> 35: ADD TEMP[5], TEMP[5], TEMP[5]
> 36: MAD TEMP[9], TEMP[3], TEMP[5], TEMP[9]
> 37: MUL TEMP[9], TEMP[1], TEMP[9]
> 38: MAD TEMP[9], TEMP[6], CONST[7], TEMP[9]
> 39: LRP OUT[0].xyz, IN[0].xxxx, CONST[8], TEMP[9]
> 40: MUL TEMP[3], TEMP[3], TEMP[5]
> 41: MUL TEMP[3], TEMP[3], TEMP[1]
> 42: DP3 TEMP[3].x, TEMP[3], IMM[1]
> 43: CMP OUT[0].w, TEMP[4].xxxx, IMM[2].xxxx, TEMP[3].xxxx
> 44: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: DP3 temp[0].x, input[7], input[7];
>  1: RSQ temp[0].x, temp[0].xxxx;
>  2: MUL temp[0], temp[0].xxxx, input[7];
>  3: DP3 temp[1].x, input[1], input[1];
>  4: RSQ temp[1].x, temp[1].xxxx;
>  5: MUL temp[1], temp[1].xxxx, input[1];
>  6: TEX temp[2], input[2], 2D[1];
>  7: MOV temp[2].x, temp[2].wwww;
>  8: MAD temp[2], temp[2], const[9].xxxx, const[9].yyyy;
>  9: MOV temp[2].z, const[10].wwww;
> 10: DP3 temp[3].x, temp[2], temp[2];
> 11: ADD temp[3].x, const[11].xxxx, -temp[3].xxxx;
> 12: RSQ temp[3].x, temp[3].xxxx;
> 13: RCP temp[2].z, temp[3].xxxx;
> 14: MOV temp[4], temp[1].zzzz;
> 15: TEX temp[5], input[5], 2D[2];
> 16: MOV temp[6], temp[5];
> 17: DP3 temp[7].x, temp[1], -temp[2];
> 18: MUL temp[7].x, temp[7].xxxx, temp[5].wwww;
> 19: MUL temp[7].x, temp[7].xxxx, const[4].wwww;
> 20: DP3 temp[8].x, temp[1], temp[2];
> 21: CMP temp[1].x, temp[4].xxxx, temp[7].xxxx, temp[8].xxxx;
> 22: TXP temp[3], input[4], 2D[3];
> 23: MUL temp[1], temp[1].xxxx, temp[3];
> 24: TEX temp[3].x, input[3], 2D[4];
> 25: MUL temp[1], temp[1], temp[3].xxxx;
> 26: MUL temp[9], temp[5], const[5];
> 27: MUL temp[3].xyz, temp[9], const[4];
> 28: CMP temp[9].xyz, temp[4].xxxx, temp[3], temp[9];
> 29: DP3 temp[0].x, temp[0], temp[2];
> 30: MAD_SAT temp[3].x, temp[0].xxxx, const[11].yyyy, -const[11].zzzz;
> 31: MUL temp[3].x, temp[3].xxxx, temp[3].xxxx;
> 32: MUL temp[3], temp[3].xxxx, const[6];
> 33: CMP temp[3], temp[4].xxxx, const[10].wwww, temp[3];
> 34: TEX temp[5], input[6], 2D[5];
> 35: ADD temp[5], temp[5], temp[5];
> 36: MAD temp[9], temp[3], temp[5], temp[9];
> 37: MUL temp[9], temp[1], temp[9];
> 38: MAD temp[9], temp[6], const[7], temp[9];
> 39: LRP output[0].xyz, input[0].xxxx, const[8], temp[9];
> 40: MUL temp[3], temp[3], temp[5];
> 41: MUL temp[3], temp[3], temp[1];
> 42: DP3 temp[3].x, temp[3], const[10];
> 43: CMP output[0].w, temp[4].xxxx, const[11].xxxx, temp[3].xxxx;
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: DP3 temp[0].x, input[7], input[7];
>  1: RSQ temp[0].x, temp[0].xxxx;
>  2: MUL temp[0], temp[0].xxxx, input[7];
>  3: DP3 temp[1].x, input[1], input[1];
>  4: RSQ temp[1].x, temp[1].xxxx;
>  5: MUL temp[1], temp[1].xxxx, input[1];
>  6: TEX temp[2], input[2], 2D[1];
>  7: MOV temp[2].x, temp[2].wwww;
>  8: MAD temp[2], temp[2], const[9].xxxx, const[9].yyyy;
>  9: MOV temp[2].z, const[10].wwww;
> 10: DP3 temp[3].x, temp[2], temp[2];
> 11: ADD temp[3].x, const[11].xxxx, -temp[3].xxxx;
> 12: RSQ temp[3].x, temp[3].xxxx;
> 13: RCP temp[2].z, temp[3].xxxx;
> 14: MOV temp[4], temp[1].zzzz;
> 15: TEX temp[5], input[5], 2D[2];
> 16: MOV temp[6], temp[5];
> 17: DP3 temp[7].x, temp[1], -temp[2];
> 18: MUL temp[7].x, temp[7].xxxx, temp[5].wwww;
> 19: MUL temp[7].x, temp[7].xxxx, const[4].wwww;
> 20: DP3 temp[8].x, temp[1], temp[2];
> 21: CMP temp[1].x, temp[4].xxxx, temp[7].xxxx, temp[8].xxxx;
> 22: TXP temp[3], input[4], 2D[3];
> 23: MUL temp[1], temp[1].xxxx, temp[3];
> 24: TEX temp[3].x, input[3], 2D[4];
> 25: MUL temp[1], temp[1], temp[3].xxxx;
> 26: MUL temp[9], temp[5], const[5];
> 27: MUL temp[3].xyz, temp[9], const[4];
> 28: CMP temp[9].xyz, temp[4].xxxx, temp[3], temp[9];
> 29: DP3 temp[0].x, temp[0], temp[2];
> 30: MAD_SAT temp[3].x, temp[0].xxxx, const[11].yyyy, -const[11].zzzz;
> 31: MUL temp[3].x, temp[3].xxxx, temp[3].xxxx;
> 32: MUL temp[3], temp[3].xxxx, const[6];
> 33: CMP temp[3], temp[4].xxxx, const[10].wwww, temp[3];
> 34: TEX temp[5], input[6], 2D[5];
> 35: ADD temp[5], temp[5], temp[5];
> 36: MAD temp[9], temp[3], temp[5], temp[9];
> 37: MUL temp[9], temp[1], temp[9];
> 38: MAD temp[9], temp[6], const[7], temp[9];
> 39: LRP output[0].xyz, input[0].xxxx, const[8], temp[9];
> 40: MUL temp[3], temp[3], temp[5];
> 41: MUL temp[3], temp[3], temp[1];
> 42: DP3 temp[3].x, temp[3], const[10];
> 43: CMP output[0].w, temp[4].xxxx, const[11].xxxx, temp[3].xxxx;
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: DP3 temp[0].x, input[7], input[7];
>  1: RSQ temp[0].x, temp[0].xxxx;
>  2: MUL temp[0], temp[0].xxxx, input[7];
>  3: DP3 temp[1].x, input[1], input[1];
>  4: RSQ temp[1].x, temp[1].xxxx;
>  5: MUL temp[1], temp[1].xxxx, input[1];
>  6: TEX temp[2], input[2], 2D[1];
>  7: MOV temp[2].x, temp[2].wwww;
>  8: MAD temp[2], temp[2], const[9].xxxx, const[9].yyyy;
>  9: MOV temp[2].z, const[10].wwww;
> 10: DP3 temp[3].x, temp[2], temp[2];
> 11: ADD temp[3].x, const[11].xxxx, -temp[3].xxxx;
> 12: RSQ temp[3].x, temp[3].xxxx;
> 13: RCP temp[2].z, temp[3].xxxx;
> 14: MOV temp[4], temp[1].zzzz;
> 15: TEX temp[5], input[5], 2D[2];
> 16: MOV temp[6], temp[5];
> 17: DP3 temp[7].x, temp[1], -temp[2];
> 18: MUL temp[7].x, temp[7].xxxx, temp[5].wwww;
> 19: MUL temp[7].x, temp[7].xxxx, const[4].wwww;
> 20: DP3 temp[8].x, temp[1], temp[2];
> 21: CMP temp[1].x, temp[4].xxxx, temp[7].xxxx, temp[8].xxxx;
> 22: TXP temp[3], input[4], 2D[3];
> 23: MUL temp[1], temp[1].xxxx, temp[3];
> 24: TEX temp[3].x, input[3], 2D[4];
> 25: MUL temp[1], temp[1], temp[3].xxxx;
> 26: MUL temp[9], temp[5], const[5];
> 27: MUL temp[3].xyz, temp[9], const[4];
> 28: CMP temp[9].xyz, temp[4].xxxx, temp[3], temp[9];
> 29: DP3 temp[0].x, temp[0], temp[2];
> 30: MAD_SAT temp[3].x, temp[0].xxxx, const[11].yyyy, -const[11].zzzz;
> 31: MUL temp[3].x, temp[3].xxxx, temp[3].xxxx;
> 32: MUL temp[3], temp[3].xxxx, const[6];
> 33: CMP temp[3], temp[4].xxxx, const[10].wwww, temp[3];
> 34: TEX temp[5], input[6], 2D[5];
> 35: ADD temp[5], temp[5], temp[5];
> 36: MAD temp[9], temp[3], temp[5], temp[9];
> 37: MUL temp[9], temp[1], temp[9];
> 38: MAD temp[9], temp[6], const[7], temp[9];
> 39: LRP output[0].xyz, input[0].xxxx, const[8], temp[9];
> 40: MUL temp[3], temp[3], temp[5];
> 41: MUL temp[3], temp[3], temp[1];
> 42: DP3 temp[3].x, temp[3], const[10];
> 43: CMP output[0].w, temp[4].xxxx, const[11].xxxx, temp[3].xxxx;
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: DP3 temp[0].x, input[7], input[7];
>  1: RSQ temp[0].x, |temp[0].xxxx|;
>  2: MUL temp[0], temp[0].xxxx, input[7];
>  3: DP3 temp[1].x, input[1], input[1];
>  4: RSQ temp[1].x, |temp[1].xxxx|;
>  5: MUL temp[1], temp[1].xxxx, input[1];
>  6: TEX temp[2], input[2], 2D[1];
>  7: MOV temp[2].x, temp[2].wwww;
>  8: MAD temp[2], temp[2], const[9].xxxx, const[9].yyyy;
>  9: MOV temp[2].z, const[10].wwww;
> 10: DP3 temp[3].x, temp[2], temp[2];
> 11: ADD temp[3].x, const[11].xxxx, -temp[3].xxxx;
> 12: RSQ temp[3].x, |temp[3].xxxx|;
> 13: RCP temp[2].z, temp[3].xxxx;
> 14: MOV temp[4], temp[1].zzzz;
> 15: TEX temp[5], input[5], 2D[2];
> 16: MOV temp[6], temp[5];
> 17: DP3 temp[7].x, temp[1], -temp[2];
> 18: MUL temp[7].x, temp[7].xxxx, temp[5].wwww;
> 19: MUL temp[7].x, temp[7].xxxx, const[4].wwww;
> 20: DP3 temp[8].x, temp[1], temp[2];
> 21: CMP temp[1].x, temp[4].xxxx, temp[7].xxxx, temp[8].xxxx;
> 22: TXP temp[3], input[4], 2D[3];
> 23: MUL temp[1], temp[1].xxxx, temp[3];
> 24: TEX temp[3].x, input[3], 2D[4];
> 25: MUL temp[1], temp[1], temp[3].xxxx;
> 26: MUL temp[9], temp[5], const[5];
> 27: MUL temp[3].xyz, temp[9], const[4];
> 28: CMP temp[9].xyz, temp[4].xxxx, temp[3], temp[9];
> 29: DP3 temp[0].x, temp[0], temp[2];
> 30: MAD_SAT temp[3].x, temp[0].xxxx, const[11].yyyy, -const[11].zzzz;
> 31: MUL temp[3].x, temp[3].xxxx, temp[3].xxxx;
> 32: MUL temp[3], temp[3].xxxx, const[6];
> 33: CMP temp[3], temp[4].xxxx, const[10].wwww, temp[3];
> 34: TEX temp[5], input[6], 2D[5];
> 35: ADD temp[5], temp[5], temp[5];
> 36: MAD temp[9], temp[3], temp[5], temp[9];
> 37: MUL temp[9], temp[1], temp[9];
> 38: MAD temp[9], temp[6], const[7], temp[9];
> 39: ADD temp[10], const[8], -temp[9];
> 40: MAD output[0].xyz, input[0].xxxx, temp[10], temp[9];
> 41: MUL temp[3], temp[3], temp[5];
> 42: MUL temp[3], temp[3], temp[1];
> 43: DP3 temp[3].x, temp[3], const[10];
> 44: CMP output[0].w, temp[4].xxxx, const[11].xxxx, temp[3].xxxx;
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: DP3 temp[0].x, input[7], input[7];
>  1: RSQ temp[0].x, |temp[0].xxxx|;
>  2: MUL temp[0], temp[0].xxxx, input[7];
>  3: DP3 temp[1].x, input[1], input[1];
>  4: RSQ temp[1].x, |temp[1].xxxx|;
>  5: MUL temp[1], temp[1].xxxx, input[1];
>  6: TEX temp[2], input[2], 2D[1];
>  7: MOV temp[2].x, temp[2].wwww;
>  8: MAD temp[2], temp[2], const[9].xxxx, const[9].yyyy;
>  9: MOV temp[2].z, const[10].wwww;
> 10: DP3 temp[3].x, temp[2], temp[2];
> 11: ADD temp[3].x, const[11].xxxx, -temp[3].xxxx;
> 12: RSQ temp[3].x, |temp[3].xxxx|;
> 13: RCP temp[2].z, temp[3].xxxx;
> 14: MOV temp[4], temp[1].zzzz;
> 15: TEX temp[5], input[5], 2D[2];
> 16: MOV temp[6], temp[5];
> 17: DP3 temp[7].x, temp[1], -temp[2];
> 18: MUL temp[7].x, temp[7].xxxx, temp[5].wwww;
> 19: MUL temp[7].x, temp[7].xxxx, const[4].wwww;
> 20: DP3 temp[8].x, temp[1], temp[2];
> 21: CMP temp[1].x, temp[4].xxxx, temp[7].xxxx, temp[8].xxxx;
> 22: TXP temp[3], input[4], 2D[3];
> 23: MUL temp[1], temp[1].xxxx, temp[3];
> 24: TEX temp[11], input[3], 2D[4];
> 25: MOV temp[3].x, temp[11];
> 26: MUL temp[1], temp[1], temp[3].xxxx;
> 27: MUL temp[9], temp[5], const[5];
> 28: MUL temp[3].xyz, temp[9], const[4];
> 29: CMP temp[9].xyz, temp[4].xxxx, temp[3], temp[9];
> 30: DP3 temp[0].x, temp[0], temp[2];
> 31: MAD_SAT temp[3].x, temp[0].xxxx, const[11].yyyy, -const[11].zzzz;
> 32: MUL temp[3].x, temp[3].xxxx, temp[3].xxxx;
> 33: MUL temp[3], temp[3].xxxx, const[6];
> 34: CMP temp[3], temp[4].xxxx, const[10].wwww, temp[3];
> 35: TEX temp[5], input[6], 2D[5];
> 36: ADD temp[5], temp[5], temp[5];
> 37: MAD temp[9], temp[3], temp[5], temp[9];
> 38: MUL temp[9], temp[1], temp[9];
> 39: MAD temp[9], temp[6], const[7], temp[9];
> 40: ADD temp[10], const[8], -temp[9];
> 41: MAD output[0].xyz, input[0].xxxx, temp[10], temp[9];
> 42: MUL temp[3], temp[3], temp[5];
> 43: MUL temp[3], temp[3], temp[1];
> 44: DP3 temp[3].x, temp[3], const[10];
> 45: CMP output[0].w, temp[4].xxxx, const[11].xxxx, temp[3].xxxx;
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: DP3 temp[0].x, input[7].xyz_, input[7].xyz_;
>  1: RSQ temp[0].x, |temp[0].x___|;
>  2: MUL temp[0].xyz, temp[0].xxx_, input[7].xyz_;
>  3: DP3 temp[1].x, input[1].xyz_, input[1].xyz_;
>  4: RSQ temp[1].x, |temp[1].x___|;
>  5: MUL temp[1].xyz, temp[1].xxx_, input[1].xyz_;
>  6: TEX temp[2].yw, input[2].xy__, 2D[1];
>  7: MOV temp[2].x, temp[2].w___;
>  8: MAD temp[2].xy, temp[2].xy__, const[9].xx__, const[9].yy__;
>  9: MOV temp[2].z, const[10].__w_;
> 10: DP3 temp[3].x, temp[2].xyz_, temp[2].xyz_;
> 11: ADD temp[3].x, const[11].x___, -temp[3].x___;
> 12: RSQ temp[3].x, |temp[3].x___|;
> 13: RCP temp[2].z, temp[3].x___;
> 14: MOV temp[4].x, temp[1].z___;
> 15: TEX temp[5], input[5].xy__, 2D[2];
> 16: MOV temp[6].xyz, temp[5].xyz_;
> 17: DP3 temp[7].x, temp[1].xyz_, -temp[2].xyz_;
> 18: MUL temp[7].x, temp[7].x___, temp[5].w___;
> 19: MUL temp[7].x, temp[7].x___, const[4].w___;
> 20: DP3 temp[8].x, temp[1].xyz_, temp[2].xyz_;
> 21: CMP temp[1].x, temp[4].x___, temp[7].x___, temp[8].x___;
> 22: TXP temp[3].xyz, input[4].xy_w, 2D[3];
> 23: MUL temp[1].xyz, temp[1].xxx_, temp[3].xyz_;
> 24: TEX temp[11].x, input[3].xy__, 2D[4];
> 25: MOV temp[3].x, temp[11].x___;
> 26: MUL temp[1].xyz, temp[1].xyz_, temp[3].xxx_;
> 27: MUL temp[9].xyz, temp[5].xyz_, const[5].xyz_;
> 28: MUL temp[3].xyz, temp[9].xyz_, const[4].xyz_;
> 29: CMP temp[9].xyz, temp[4].xxx_, temp[3].xyz_, temp[9].xyz_;
> 30: DP3 temp[0].x, temp[0].xyz_, temp[2].xyz_;
> 31: MAD_SAT temp[3].x, temp[0].x___, const[11].y___, -const[11].z___;
> 32: MUL temp[3].x, temp[3].x___, temp[3].x___;
> 33: MUL temp[3].xyz, temp[3].xxx_, const[6].xyz_;
> 34: CMP temp[3].xyz, temp[4].xxx_, const[10].www_, temp[3].xyz_;
> 35: TEX temp[5].xyz, input[6].xy__, 2D[5];
> 36: ADD temp[5].xyz, temp[5].xyz_, temp[5].xyz_;
> 37: MAD temp[9].xyz, temp[3].xyz_, temp[5].xyz_, temp[9].xyz_;
> 38: MUL temp[9].xyz, temp[1].xyz_, temp[9].xyz_;
> 39: MAD temp[9].xyz, temp[6].xyz_, const[7].xyz_, temp[9].xyz_;
> 40: ADD temp[10].xyz, const[8].xyz_, -temp[9].xyz_;
> 41: MAD output[0].xyz, input[0].xxx_, temp[10].xyz_, temp[9].xyz_;
> 42: MUL temp[3].xyz, temp[3].xyz_, temp[5].xyz_;
> 43: MUL temp[3].xyz, temp[3].xyz_, temp[1].xyz_;
> 44: DP3 temp[3].x, temp[3].xyz_, const[10].xyz_;
> 45: CMP output[0].w, temp[4].___x, const[11].___x, temp[3].___x;
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: DP3 temp[0].x, input[7].xyz_, input[7].xyz_;
>  1: RSQ temp[0].x, |temp[0].x___|;
>  2: MUL temp[0].xyz, temp[0].xxx_, input[7].xyz_;
>  3: DP3 temp[1].x, input[1].xyz_, input[1].xyz_;
>  4: RSQ temp[1].x, |temp[1].x___|;
>  5: MUL temp[1].xyz, temp[1].xxx_, input[1].xyz_;
>  6: TEX temp[2].yw, input[2].xy__, 2D[1];
>  7: MOV temp[2].x, temp[2].w___;
>  8: MAD temp[2].xy, temp[2].xy__, const[9].xx__, none.-1-1__;
>  9: MOV temp[2].z, none.__0_;
> 10: DP3 temp[3].x, temp[2].xyz_, temp[2].xyz_;
> 11: ADD temp[3].x, none.1___, -temp[3].x___;
> 12: RSQ temp[3].x, |temp[3].x___|;
> 13: RCP temp[2].z, temp[3].x___;
> 14: MOV temp[4].x, temp[1].z___;
> 15: TEX temp[5], input[5].xy__, 2D[2];
> 16: MOV temp[6].xyz, temp[5].xyz_;
> 17: DP3 temp[7].x, temp[1].xyz_, -temp[2].xyz_;
> 18: MUL temp[7].x, temp[7].x___, temp[5].w___;
> 19: MUL temp[7].x, temp[7].x___, const[4].w___;
> 20: DP3 temp[8].x, temp[1].xyz_, temp[2].xyz_;
> 21: CMP temp[1].x, temp[4].x___, temp[7].x___, temp[8].x___;
> 22: TXP temp[3].xyz, input[4].xy_w, 2D[3];
> 23: MUL temp[1].xyz, temp[1].xxx_, temp[3].xyz_;
> 24: TEX temp[11].x, input[3].xy__, 2D[4];
> 25: MUL temp[1].xyz, temp[1].xyz_, temp[11].xxx_;
> 26: MUL temp[9].xyz, temp[5].xyz_, const[5].xyz_;
> 27: MUL temp[3].xyz, temp[9].xyz_, const[4].xyz_;
> 28: CMP temp[9].xyz, temp[4].xxx_, temp[3].xyz_, temp[9].xyz_;
> 29: DP3 temp[0].x, temp[0].xyz_, temp[2].xyz_;
> 30: MAD_SAT temp[3].x, temp[0].x___, const[11].y___, -const[11].z___;
> 31: MUL temp[3].x, temp[3].x___, temp[3].x___;
> 32: MUL temp[3].xyz, temp[3].xxx_, const[6].xyz_;
> 33: CMP temp[3].xyz, temp[4].xxx_, none.000_, temp[3].xyz_;
> 34: TEX temp[5].xyz, input[6].xy__, 2D[5];
> 35: ADD temp[5].xyz, temp[5].xyz_, temp[5].xyz_;
> 36: MAD temp[9].xyz, temp[3].xyz_, temp[5].xyz_, temp[9].xyz_;
> 37: MUL temp[9].xyz, temp[1].xyz_, temp[9].xyz_;
> 38: MAD temp[9].xyz, temp[6].xyz_, const[7].xyz_, temp[9].xyz_;
> 39: ADD temp[10].xyz, const[8].xyz_, -temp[9].xyz_;
> 40: MAD output[0].xyz, input[0].xxx_, temp[10].xyz_, temp[9].xyz_;
> 41: MUL temp[3].xyz, temp[3].xyz_, temp[5].xyz_;
> 42: MUL temp[3].xyz, temp[3].xyz_, temp[1].xyz_;
> 43: DP3 temp[3].x, temp[3].xyz_, const[10].xyz_;
> 44: CMP output[0].w, temp[4].___x, none.___1, temp[3].___x;
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: DP3 temp[0].x, input[7].xyz_, input[7].xyz_;
>  1: RSQ temp[0].x, |temp[0].x___|;
>  2: MUL temp[0].xyz, temp[0].xxx_, input[7].xyz_;
>  3: DP3 temp[1].x, input[1].xyz_, input[1].xyz_;
>  4: RSQ temp[1].x, |temp[1].x___|;
>  5: MUL temp[1].xyz, temp[1].xxx_, input[1].xyz_;
>  6: TEX temp[2].yw, input[2].xy__, 2D[1];
>  7: MOV temp[2].x, temp[2].w___;
>  8: MAD temp[2].xy, temp[2].xy__, const[9].xx__, none.-1-1__;
>  9: MOV temp[2].z, none.__0_;
> 10: DP3 temp[3].x, temp[2].xyz_, temp[2].xyz_;
> 11: ADD temp[3].x, none.1___, -temp[3].x___;
> 12: RSQ temp[3].x, |temp[3].x___|;
> 13: RCP temp[2].z, temp[3].x___;
> 14: MOV temp[4].x, temp[1].z___;
> 15: TEX temp[5], input[5].xy__, 2D[2];
> 16: MOV temp[6].xyz, temp[5].xyz_;
> 17: DP3 temp[7].x, temp[1].xyz_, -temp[2].xyz_;
> 18: MUL temp[7].x, temp[7].x___, temp[5].w___;
> 19: MUL temp[7].x, temp[7].x___, const[4].w___;
> 20: DP3 temp[8].x, temp[1].xyz_, temp[2].xyz_;
> 21: CMP temp[1].x, temp[4].x___, temp[7].x___, temp[8].x___;
> 22: TXP temp[3].xyz, input[4].xy_w, 2D[3];
> 23: MUL temp[1].xyz, temp[1].xxx_, temp[3].xyz_;
> 24: TEX temp[11].x, input[3].xy__, 2D[4];
> 25: MUL temp[1].xyz, temp[1].xyz_, temp[11].xxx_;
> 26: MUL temp[9].xyz, temp[5].xyz_, const[5].xyz_;
> 27: MUL temp[3].xyz, temp[9].xyz_, const[4].xyz_;
> 28: CMP temp[9].xyz, temp[4].xxx_, temp[3].xyz_, temp[9].xyz_;
> 29: DP3 temp[0].x, temp[0].xyz_, temp[2].xyz_;
> 30: MAD_SAT temp[3].x, temp[0].x___, const[11].y___, -const[11].z___;
> 31: MUL temp[3].x, temp[3].x___, temp[3].x___;
> 32: MUL temp[3].xyz, temp[3].xxx_, const[6].xyz_;
> 33: CMP temp[3].xyz, temp[4].xxx_, none.000_, temp[3].xyz_;
> 34: TEX temp[5].xyz, input[6].xy__, 2D[5];
> 35: ADD temp[5].xyz, temp[5].xyz_, temp[5].xyz_;
> 36: MAD temp[9].xyz, temp[3].xyz_, temp[5].xyz_, temp[9].xyz_;
> 37: MUL temp[9].xyz, temp[1].xyz_, temp[9].xyz_;
> 38: MAD temp[9].xyz, temp[6].xyz_, const[7].xyz_, temp[9].xyz_;
> 39: ADD temp[10].xyz, const[8].xyz_, -temp[9].xyz_;
> 40: MAD output[0].xyz, input[0].xxx_, temp[10].xyz_, temp[9].xyz_;
> 41: MUL temp[3].xyz, temp[3].xyz_, temp[5].xyz_;
> 42: MUL temp[3].xyz, temp[3].xyz_, temp[1].xyz_;
> 43: DP3 temp[3].x, temp[3].xyz_, const[10].xyz_;
> 44: CMP output[0].w, temp[4].___x, none.___1, temp[3].___x;
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: src0.xyz = input[7]
>     DP3 temp[0].x, src0.xyz, src0.xyz
>  1: src0.xyz = temp[0]
>     REPL_ALPHA temp[0].x
>     RSQ, |src0.x|
>  2: src0.xyz = temp[0], src1.xyz = input[7]
>     MAD temp[0].xyz, src0.xxx, src1.xyz, src0.000
>  3: src0.xyz = input[1]
>     DP3 temp[1].x, src0.xyz, src0.xyz
>  4: src0.xyz = temp[1]
>     REPL_ALPHA temp[1].x
>     RSQ, |src0.x|
>  5: src0.xyz = temp[1], src1.xyz = input[1]
>     MAD temp[1].xyz, src0.xxx, src1.xyz, src0.000
>  6: TEX temp[2].yw, input[2].xy__, 2D[1];
>  7: src0.w = temp[2]
>     MAD temp[2].x, src0.w__, src0.111, src0.000
>  8: src0.xyz = temp[2], src1.xyz = const[9]
>     MAD temp[2].xy, src0.xy_, src1.xx_, -src0.11_
>  9: 
>     MAD temp[2].z, src0.__0, src0.111, src0.000
> 10: src0.xyz = temp[2]
>     DP3 temp[3].x, src0.xyz, src0.xyz
> 11: src0.xyz = temp[3]
>     MAD temp[3].x, src0.1__, src0.111, -src0.x__
> 12: src0.xyz = temp[3]
>     REPL_ALPHA temp[3].x
>     RSQ, |src0.x|
> 13: src0.xyz = temp[3]
>     REPL_ALPHA temp[2].z
>     RCP, src0.x
> 14: src0.xyz = temp[1]
>     MAD temp[4].x, src0.z__, src0.111, src0.000
> 15: TEX temp[5], input[5].xy__, 2D[2];
> 16: src0.xyz = temp[5]
>     MAD temp[6].xyz, src0.xyz, src0.111, src0.000
> 17: src0.xyz = temp[1], src1.xyz = temp[2]
>     DP3 temp[7].x, src0.xyz, -src1.xyz
> 18: src0.xyz = temp[7], src0.w = temp[5]
>     MAD temp[7].x, src0.x__, src0.w__, src0.000
> 19: src0.xyz = temp[7], src0.w = const[4]
>     MAD temp[7].x, src0.x__, src0.w__, src0.000
> 20: src0.xyz = temp[1], src1.xyz = temp[2]
>     DP3 temp[8].x, src0.xyz, src1.xyz
> 21: src0.xyz = temp[8], src1.xyz = temp[7], src2.xyz = temp[4]
>     CMP temp[1].x, src0.x__, src1.x__, src2.x__
> 22: TXP temp[3].xyz, input[4].xy_w, 2D[3];
> 23: src0.xyz = temp[1], src1.xyz = temp[3]
>     MAD temp[1].xyz, src0.xxx, src1.xyz, src0.000
> 24: TEX temp[11].x, input[3].xy__, 2D[4];
> 25: src0.xyz = temp[1], src1.xyz = temp[11]
>     MAD temp[1].xyz, src0.xyz, src1.xxx, src0.000
> 26: src0.xyz = temp[5], src1.xyz = const[5]
>     MAD temp[9].xyz, src0.xyz, src1.xyz, src0.000
> 27: src0.xyz = temp[9], src1.xyz = const[4]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src0.000
> 28: src0.xyz = temp[9], src1.xyz = temp[3], src2.xyz = temp[4]
>     CMP temp[9].xyz, src0.xyz, src1.xyz, src2.xxx
> 29: src0.xyz = temp[0], src1.xyz = temp[2]
>     DP3 temp[0].x, src0.xyz, src1.xyz
> 30: src0.xyz = temp[0], src1.xyz = const[11]
>     MAD_SAT temp[3].x, src0.x__, src1.y__, -src1.z__
> 31: src0.xyz = temp[3]
>     MAD temp[3].x, src0.x__, src0.x__, src0.000
> 32: src0.xyz = temp[3], src1.xyz = const[6]
>     MAD temp[3].xyz, src0.xxx, src1.xyz, src0.000
> 33: src0.xyz = temp[3], src1.xyz = temp[4]
>     CMP temp[3].xyz, src0.xyz, src0.000, src1.xxx
> 34: TEX temp[5].xyz, input[6].xy__, 2D[5];
> 35: src0.xyz = temp[5]
>     MAD temp[5].xyz, src0.xyz, src0.111, src0.xyz
> 36: src0.xyz = temp[3], src1.xyz = temp[5], src2.xyz = temp[9]
>     MAD temp[9].xyz, src0.xyz, src1.xyz, src2.xyz
> 37: src0.xyz = temp[1], src1.xyz = temp[9]
>     MAD temp[9].xyz, src0.xyz, src1.xyz, src0.000
> 38: src0.xyz = temp[6], src1.xyz = const[7], src2.xyz = temp[9]
>     MAD temp[9].xyz, src0.xyz, src1.xyz, src2.xyz
> 39: src0.xyz = const[8], src1.xyz = temp[9]
>     MAD temp[10].xyz, src0.xyz, src0.111, -src1.xyz
> 40: src0.xyz = input[0], src1.xyz = temp[10], src2.xyz = temp[9]
>     MAD color[0].xyz, src0.xxx, src1.xyz, src2.xyz
> 41: src0.xyz = temp[3], src1.xyz = temp[5]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src0.000
> 42: src0.xyz = temp[3], src1.xyz = temp[1]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src0.000
> 43: src0.xyz = temp[3], src1.xyz = const[10]
>     DP3 temp[3].x, src0.xyz, src1.xyz
> 44: src0.xyz = temp[3], src1.xyz = temp[4]
>     CMP color[0].w, src0.x, src0.1, src1.x
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[11].x, input[3].xy__, 2D[4];
>  2: TEX temp[5], input[5].xy__, 2D[2];
>  3: TEX temp[2].yw, input[2].xy__, 2D[1];
>  4: src0.w = temp[2]
>     MAD temp[2].x, src0.w__, src0.111, src0.000
>  5: src0.xyz = temp[2], src1.xyz = const[9]
>     MAD temp[2].xy, src0.xy_, src1.xx_, -src0.11_
>  6: src0.xyz = temp[5]
>     MAD temp[6].xyz, src0.xyz, src0.111, src0.000
>  7: src0.xyz = temp[5], src1.xyz = const[5]
>     MAD temp[9].xyz, src0.xyz, src1.xyz, src0.000
>  8: 
>     MAD temp[2].z, src0.__0, src0.111, src0.000
>  9: src0.xyz = temp[2]
>     DP3 temp[3].x, src0.xyz, src0.xyz
> 10: src0.xyz = temp[3]
>     MAD temp[3].x, src0.1__, src0.111, -src0.x__
> 11: src0.xyz = temp[3]
>     REPL_ALPHA temp[3].x
>     RSQ, |src0.x|
> 12: src0.xyz = temp[3]
>     REPL_ALPHA temp[2].z
>     RCP, src0.x
> 13: src0.xyz = input[1]
>     DP3 temp[1].x, src0.xyz, src0.xyz
> 14: src0.xyz = temp[1]
>     REPL_ALPHA temp[1].x
>     RSQ, |src0.x|
> 15: src0.xyz = temp[1], src1.xyz = input[1]
>     MAD temp[1].xyz, src0.xxx, src1.xyz, src0.000
> 16: src0.xyz = temp[1]
>     MAD temp[4].x, src0.z__, src0.111, src0.000
> 17: src0.xyz = temp[1], src1.xyz = temp[2]
>     DP3 temp[7].x, src0.xyz, -src1.xyz
> 18: src0.xyz = temp[7], src0.w = temp[5]
>     MAD temp[7].x, src0.x__, src0.w__, src0.000
> 19: src0.xyz = temp[7], src0.w = const[4]
>     MAD temp[7].x, src0.x__, src0.w__, src0.000
> 20: src0.xyz = temp[1], src1.xyz = temp[2]
>     DP3 temp[8].x, src0.xyz, src1.xyz
> 21: src0.xyz = temp[8], src1.xyz = temp[7], src2.xyz = temp[4]
>     CMP temp[1].x, src0.x__, src1.x__, src2.x__
> 22: src0.xyz = input[7]
>     DP3 temp[0].x, src0.xyz, src0.xyz
> 23: src0.xyz = temp[0]
>     REPL_ALPHA temp[0].x
>     RSQ, |src0.x|
> 24: src0.xyz = temp[0], src1.xyz = input[7]
>     MAD temp[0].xyz, src0.xxx, src1.xyz, src0.000
> 25: src0.xyz = temp[0], src1.xyz = temp[2]
>     DP3 temp[0].x, src0.xyz, src1.xyz
> 26: BEGIN_TEX;
> 27: TXP temp[3].xyz, input[4].xy_w, 2D[3];
> 28: TEX temp[5].xyz, input[6].xy__, 2D[5];
> 29: src0.xyz = temp[5]
>     MAD temp[5].xyz, src0.xyz, src0.111, src0.xyz
> 30: src0.xyz = temp[1], src1.xyz = temp[3]
>     MAD temp[1].xyz, src0.xxx, src1.xyz, src0.000
> 31: src0.xyz = temp[1], src1.xyz = temp[11]
>     MAD temp[1].xyz, src0.xyz, src1.xxx, src0.000
> 32: src0.xyz = temp[9], src1.xyz = const[4]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src0.000
> 33: src0.xyz = temp[9], src1.xyz = temp[3], src2.xyz = temp[4]
>     CMP temp[9].xyz, src0.xyz, src1.xyz, src2.xxx
> 34: src0.xyz = temp[0], src1.xyz = const[11]
>     MAD_SAT temp[3].x, src0.x__, src1.y__, -src1.z__
> 35: src0.xyz = temp[3]
>     MAD temp[3].x, src0.x__, src0.x__, src0.000
> 36: src0.xyz = temp[3], src1.xyz = const[6]
>     MAD temp[3].xyz, src0.xxx, src1.xyz, src0.000
> 37: src0.xyz = temp[3], src1.xyz = temp[4]
>     CMP temp[3].xyz, src0.xyz, src0.000, src1.xxx
> 38: src0.xyz = temp[3], src1.xyz = temp[5], src2.xyz = temp[9]
>     MAD temp[9].xyz, src0.xyz, src1.xyz, src2.xyz
> 39: src0.xyz = temp[1], src1.xyz = temp[9]
>     MAD temp[9].xyz, src0.xyz, src1.xyz, src0.000
> 40: src0.xyz = temp[6], src1.xyz = const[7], src2.xyz = temp[9]
>     MAD temp[9].xyz, src0.xyz, src1.xyz, src2.xyz
> 41: src0.xyz = const[8], src1.xyz = temp[9]
>     MAD temp[10].xyz, src0.xyz, src0.111, -src1.xyz
> 42: src0.xyz = input[0], src1.xyz = temp[10], src2.xyz = temp[9]
>     MAD color[0].xyz, src0.xxx, src1.xyz, src2.xyz
> 43: src0.xyz = temp[3], src1.xyz = temp[5]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src0.000
> 44: src0.xyz = temp[3], src1.xyz = temp[1]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src0.000
> 45: src0.xyz = temp[3], src1.xyz = const[10]
>     DP3 temp[3].x, src0.xyz, src1.xyz
> 46: src0.xyz = temp[3], src1.xyz = temp[4]
>     CMP color[0].w, src0.x, src0.1, src1.x
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[13].x, temp[3].xy__, 2D[4];
>  2: TEX temp[9], temp[5].xy__, 2D[2];
>  3: TEX temp[3].yw, temp[2].xy__, 2D[1];
>  4: src0.w = temp[3]
>     MAD temp[3].x, src0.w__, src0.111, src0.000
>  5: src0.xyz = temp[3], src1.xyz = const[9]
>     MAD temp[3].xy, src0.xy_, src1.xx_, -src0.11_
>  6: src0.xyz = temp[9]
>     MAD temp[10].xyz, src0.xyz, src0.111, src0.000
>  7: src0.xyz = temp[9], src1.xyz = const[5]
>     MAD temp[12].xyz, src0.xyz, src1.xyz, src0.000
>  8: 
>     MAD temp[3].z, src0.__0, src0.111, src0.000
>  9: src0.xyz = temp[3]
>     DP3 temp[5].x, src0.xyz, src0.xyz
> 10: src0.xyz = temp[5]
>     MAD temp[5].x, src0.1__, src0.111, -src0.x__
> 11: src0.xyz = temp[5]
>     REPL_ALPHA temp[5].x
>     RSQ, |src0.x|
> 12: src0.xyz = temp[5]
>     REPL_ALPHA temp[3].z
>     RCP, src0.x
> 13: src0.xyz = temp[1]
>     DP3 temp[2].x, src0.xyz, src0.xyz
> 14: src0.xyz = temp[2]
>     REPL_ALPHA temp[2].x
>     RSQ, |src0.x|
> 15: src0.xyz = temp[2], src1.xyz = temp[1]
>     MAD temp[2].xyz, src0.xxx, src1.xyz, src0.000
> 16: src0.xyz = temp[2]
>     MAD temp[8].x, src0.z__, src0.111, src0.000
> 17: src0.xyz = temp[2], src1.xyz = temp[3]
>     DP3 temp[1].x, src0.xyz, -src1.xyz
> 18: src0.xyz = temp[1], src0.w = temp[9]
>     MAD temp[1].x, src0.x__, src0.w__, src0.000
> 19: src0.xyz = temp[1], src0.w = const[4]
>     MAD temp[1].x, src0.x__, src0.w__, src0.000
> 20: src0.xyz = temp[2], src1.xyz = temp[3]
>     DP3 temp[11].x, src0.xyz, src1.xyz
> 21: src0.xyz = temp[11], src1.xyz = temp[1], src2.xyz = temp[8]
>     CMP temp[2].x, src0.x__, src1.x__, src2.x__
> 22: src0.xyz = temp[7]
>     DP3 temp[1].x, src0.xyz, src0.xyz
> 23: src0.xyz = temp[1]
>     REPL_ALPHA temp[1].x
>     RSQ, |src0.x|
> 24: src0.xyz = temp[1], src1.xyz = temp[7]
>     MAD temp[1].xyz, src0.xxx, src1.xyz, src0.000
> 25: src0.xyz = temp[1], src1.xyz = temp[3]
>     DP3 temp[1].x, src0.xyz, src1.xyz
> 26: BEGIN_TEX;
> 27: TXP temp[5].xyz, temp[4].xy_w, 2D[3];
> 28: TEX temp[9].xyz, temp[6].xy__, 2D[5];
> 29: src0.xyz = temp[9]
>     MAD temp[9].xyz, src0.xyz, src0.111, src0.xyz
> 30: src0.xyz = temp[2], src1.xyz = temp[5]
>     MAD temp[2].xyz, src0.xxx, src1.xyz, src0.000
> 31: src0.xyz = temp[2], src1.xyz = temp[13]
>     MAD temp[2].xyz, src0.xyz, src1.xxx, src0.000
> 32: src0.xyz = temp[12], src1.xyz = const[4]
>     MAD temp[5].xyz, src0.xyz, src1.xyz, src0.000
> 33: src0.xyz = temp[12], src1.xyz = temp[5], src2.xyz = temp[8]
>     CMP temp[12].xyz, src0.xyz, src1.xyz, src2.xxx
> 34: src0.xyz = temp[1], src1.xyz = const[11]
>     MAD_SAT temp[5].x, src0.x__, src1.y__, -src1.z__
> 35: src0.xyz = temp[5]
>     MAD temp[5].x, src0.x__, src0.x__, src0.000
> 36: src0.xyz = temp[5], src1.xyz = const[6]
>     MAD temp[5].xyz, src0.xxx, src1.xyz, src0.000
> 37: src0.xyz = temp[5], src1.xyz = temp[8]
>     CMP temp[5].xyz, src0.xyz, src0.000, src1.xxx
> 38: src0.xyz = temp[5], src1.xyz = temp[9], src2.xyz = temp[12]
>     MAD temp[12].xyz, src0.xyz, src1.xyz, src2.xyz
> 39: src0.xyz = temp[2], src1.xyz = temp[12]
>     MAD temp[12].xyz, src0.xyz, src1.xyz, src0.000
> 40: src0.xyz = temp[10], src1.xyz = const[7], src2.xyz = temp[12]
>     MAD temp[12].xyz, src0.xyz, src1.xyz, src2.xyz
> 41: src0.xyz = const[8], src1.xyz = temp[12]
>     MAD temp[1].xyz, src0.xyz, src0.111, -src1.xyz
> 42: src0.xyz = temp[0], src1.xyz = temp[1], src2.xyz = temp[12]
>     MAD color[0].xyz, src0.xxx, src1.xyz, src2.xyz
> 43: src0.xyz = temp[5], src1.xyz = temp[9]
>     MAD temp[5].xyz, src0.xyz, src1.xyz, src0.000
> 44: src0.xyz = temp[5], src1.xyz = temp[2]
>     MAD temp[5].xyz, src0.xyz, src1.xyz, src0.000
> 45: src0.xyz = temp[5], src1.xyz = const[10]
>     DP3 temp[5].x, src0.xyz, src1.xyz
> 46: src0.xyz = temp[5], src1.xyz = temp[8]
>     CMP color[0].w, src0.x, src0.1, src1.x
>pc=50*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 21, tex_end: 2  (code_addr: 00040540)
>  TEX:
>    TEX t13, t3, texture[4]   (0000a343)
>    TEX t9, t5, texture[2]   (00009245)
>    TEX t3, t2, texture[1]   (000088c2)
>  0: xyz:  t0  t0  t0 -> t3.x                 (008c0000)
>       w:  t3  t0  t0 ->                      (00000003)
>     xyz:   t3.www      1.0      0.0    op: 00050a8c
>       w:     t0.x     t0.x     t0.x    op: 00000000
>  1: xyz:  t3  c9  t0 -> t3.xy                (018c0a43)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t3.xyz   c9.xxx     -1.0    op: 000d4280
>       w:     t3.x     t3.x     t3.x    op: 00000000
>  2: xyz:  t9  t0  t0 -> t10.xyz              (03a80009)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t9.xyz      1.0      0.0    op: 00050a80
>       w:     t9.x     t9.x     t9.x    op: 00000000
>  3: xyz:  t9  c5  t0 -> t12.xyz              (03b00949)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t9.xyz   c5.xyz      0.0    op: 00050200
>       w:     t9.x     t9.x     t9.x    op: 00000000
>  4: xyz:  t0  t0  t0 -> t3.z                 (020c0000)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:      0.0      1.0      0.0    op: 00050a94
>       w:     t0.x     t0.x     t0.x    op: 00000000
>  5: xyz:  t3  t0  t0 -> t5.x                 (00940003)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t3.xyz   t3.xyz   t3.xxx    op: 00804000
>       w:     t3.x     t3.x     t3.x    op: 00000000
>  6: xyz:  t5  t0  t0 -> t5.x                 (00940005)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:      1.0      1.0  -t5.xyz    op: 00080a95
>       w:     t5.x     t5.x     t5.x    op: 00000000
>  7: xyz:  t5  t0  t0 -> t5.x                 (00940005)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t5.xxx   t5.xxx   t5.xxx    op: 05004081
>       w:   |t5.x|     t5.x     t5.x    op: 05800040
>  8: xyz:  t5  t0  t0 -> t3.z                 (020c0005)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t5.xxx   t5.xxx   t5.xxx    op: 05004081
>       w:     t5.x     t5.x     t5.x    op: 05000000
>  9: xyz:  t1  t0  t0 -> t2.x                 (00880001)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   t1.xyz   t1.xxx    op: 00804000
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 10: xyz:  t2  t0  t0 -> t2.x                 (00880002)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xxx   t2.xxx   t2.xxx    op: 05004081
>       w:   |t2.x|     t2.x     t2.x    op: 05800040
> 11: xyz:  t2  t1  t0 -> t2.xyz               (03880042)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xxx   t1.xyz      0.0    op: 00050201
>       w:     t2.x     t2.x     t2.x    op: 00000000
> 12: xyz:  t2  t0  t0 -> t8.x                 (00a00002)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.zzz      1.0      0.0    op: 00050a83
>       w:     t2.x     t2.x     t2.x    op: 00000000
> 13: xyz:  t2  t3  t0 -> t1.x                 (008400c2)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz  -t3.xyz   t2.xxx    op: 00805200
>       w:     t2.x     t2.x     t2.x    op: 00000000
> 14: xyz:  t1  t0  t0 -> t1.x                 (00840001)
>       w:  t9  t0  t0 ->                      (00000009)
>     xyz:   t1.xyz   t9.www      0.0    op: 00050600
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 15: xyz:  t1  t0  t0 -> t1.x                 (00840001)
>       w:  c4  t0  t0 ->                      (00000024)
>     xyz:   t1.xyz   c4.www      0.0    op: 00050600
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 16: xyz:  t2  t3  t0 -> t11.x                (00ac00c2)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz   t3.xyz   t2.xxx    op: 00804200
>       w:     t2.x     t2.x     t2.x    op: 00000000
> 17: xyz: t11  t1  t8 -> t2.x                 (0088804b)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:  t11.xyz   t1.xyz   t8.xyz    op: 04020200
>       w:    t11.x    t11.x    t11.x    op: 00000000
> 18: xyz:  t7  t0  t0 -> t1.x                 (00840007)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t7.xyz   t7.xyz   t7.xxx    op: 00804000
>       w:     t7.x     t7.x     t7.x    op: 00000000
> 19: xyz:  t1  t0  t0 -> t1.x                 (00840001)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xxx   t1.xxx   t1.xxx    op: 05004081
>       w:   |t1.x|     t1.x     t1.x    op: 05800040
> 20: xyz:  t1  t7  t0 -> t1.xyz               (038401c1)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xxx   t7.xyz      0.0    op: 00050201
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 21: xyz:  t1  t3  t0 -> t1.x                 (008400c1)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   t3.xyz   t1.xxx    op: 00804200
>       w:     t1.x     t1.x     t1.x    op: 00000000
>NODE 1: alu_offset: 22, tex_offset: 3, alu_end: 17, tex_end: 1  (code_addr: 00423456)
>  TEX:
>    TXP t5, t4, texture[3]   (00019944)
>    TEX t9, t6, texture[5]   (0000aa46)
> 22: xyz:  t9  t0  t0 -> t9.xyz               (03a40009)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t9.xyz      1.0   t9.xyz    op: 00000a80
>       w:     t9.x     t9.x     t9.x    op: 00000000
> 23: xyz:  t2  t5  t0 -> t2.xyz               (03880142)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xxx   t5.xyz      0.0    op: 00050201
>       w:     t2.x     t2.x     t2.x    op: 00000000
> 24: xyz:  t2 t13  t0 -> t2.xyz               (03880342)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz  t13.xxx      0.0    op: 00050280
>       w:     t2.x     t2.x     t2.x    op: 00000000
> 25: xyz: t12  c4  t0 -> t5.xyz               (0394090c)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:  t12.xyz   c4.xyz      0.0    op: 00050200
>       w:    t12.x    t12.x    t12.x    op: 00000000
> 26: xyz: t12  t5  t8 -> t12.xyz              (03b0814c)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:  t12.xyz   t5.xyz   t8.xxx    op: 04024200
>       w:    t12.x    t12.x    t12.x    op: 00000000
> 27: xyz:  t1 c11  t0 -> t5.x                 (00940ac1)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz  c11.yyy -c11.zzz    op: 4009c300
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 28: xyz:  t5  t0  t0 -> t5.x                 (00940005)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t5.xyz   t5.xyz      0.0    op: 00050000
>       w:     t5.x     t5.x     t5.x    op: 00000000
> 29: xyz:  t5  c6  t0 -> t5.xyz               (03940985)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t5.xxx   c6.xyz      0.0    op: 00050201
>       w:     t5.x     t5.x     t5.x    op: 00000000
> 30: xyz:  t5  t8  t0 -> t5.xyz               (03940205)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t5.xyz      0.0   t8.xxx    op: 04014a00
>       w:     t5.x     t5.x     t5.x    op: 00000000
> 31: xyz:  t5  t9 t12 -> t12.xyz              (03b0c245)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t5.xyz   t9.xyz  t12.xyz    op: 00020200
>       w:     t5.x     t5.x     t5.x    op: 00000000
> 32: xyz:  t2 t12  t0 -> t12.xyz              (03b00302)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz  t12.xyz      0.0    op: 00050200
>       w:     t2.x     t2.x     t2.x    op: 00000000
> 33: xyz: t10  c7 t12 -> t12.xyz              (03b0c9ca)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:  t10.xyz   c7.xyz  t12.xyz    op: 00020200
>       w:    t10.x    t10.x    t10.x    op: 00000000
> 34: xyz:  c8 t12  t0 -> t1.xyz               (03840328)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   c8.xyz      1.0 -t12.xyz    op: 00090a80
>       w:     c8.x     c8.x     c8.x    op: 00000000
> 35: xyz:  t0  t1 t12 -> o0.xyz               (1c00c040)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xxx   t1.xyz  t12.xyz    op: 00020201
>       w:     t0.x     t0.x     t0.x    op: 00000000
> 36: xyz:  t5  t9  t0 -> t5.xyz               (03940245)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t5.xyz   t9.xyz      0.0    op: 00050200
>       w:     t5.x     t5.x     t5.x    op: 00000000
> 37: xyz:  t5  t2  t0 -> t5.xyz               (03940085)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t5.xyz   t2.xyz      0.0    op: 00050200
>       w:     t5.x     t5.x     t5.x    op: 00000000
> 38: xyz:  t5 c10  t0 -> t5.x                 (00940a85)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t5.xyz  c10.xyz   t5.xxx    op: 00804200
>       w:     t5.x     t5.x     t5.x    op: 00000000
> 39: xyz:  t5  t8  t0 ->                      (00000205)
>       w:  t0  t0  t0 -> o0.w                 (01000000)
>     xyz:   t5.xxx   t5.xxx   t5.xxx    op: 00004081
>       w:     t5.x      1.0     t8.x    op: 0300c880
>r300: Initial fragment program
>FRAG
>DCL IN[0], POSITION, LINEAR
>DCL IN[1], COLOR, LINEAR
>DCL IN[2], GENERIC[0], PERSPECTIVE
>DCL IN[3], GENERIC[2], PERSPECTIVE
>DCL IN[4], GENERIC[3], PERSPECTIVE
>DCL IN[5], GENERIC[4], PERSPECTIVE
>DCL IN[6], GENERIC[5], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL SAMP[1]
>DCL SAMP[2]
>DCL SAMP[3]
>DCL SAMP[4]
>DCL CONST[7..8]
>DCL CONST[1..6]
>DCL TEMP[0..9]
>IMM FLT32 {    2.0000,     1.0000,     0.0000,     0.0000}
>  0: MOV TEMP[0], IN[0]
>  1: SUB TEMP[0].y, CONST[8], IN[0]
>  2: TEX TEMP[1], IN[2], SAMP[0], 2D
>  3: MOV TEMP[1].x, TEMP[1].wwww
>  4: MAD TEMP[2], TEMP[1], IMM[0].xxxx, -IMM[0].yyyy
>  5: MUL TEMP[2].xy, TEMP[2], CONST[1].zzzz
>  6: MOV TEMP[2].z, IMM[0].zzzz
>  7: DP3 TEMP[3].x, TEMP[2], TEMP[2]
>  8: ADD TEMP[3].x, IMM[0].yyyy, -TEMP[3].xxxx
>  9: RSQ TEMP[3].x, TEMP[3].xxxx
> 10: RCP TEMP[2].z, TEMP[3].xxxx
> 11: TEX TEMP[4], IN[2], SAMP[1], 2D
> 12: DP3 TEMP[3].x, TEMP[2], IN[4]
> 13: DP3 TEMP[3].y, TEMP[2], IN[5]
> 14: DP3 TEMP[3].z, TEMP[2], IN[6]
> 15: MUL TEMP[5], CONST[1], CONST[2]
> 16: MAD TEMP[5], TEMP[5], TEMP[3], TEMP[0]
> 17: DP3 TEMP[6].w, IN[3], IN[3]
> 18: RSQ TEMP[6].w, TEMP[6].wwww
> 19: MUL TEMP[6], IN[3], TEMP[6].wwww
> 20: DP3 TEMP[2], TEMP[6], TEMP[3]
> 21: MUL TEMP[2], TEMP[2], IMM[0].xxxx
> 22: MAD TEMP[7], -TEMP[3], TEMP[2].xxxx, TEMP[6]
> 23: DP3_SAT TEMP[1], TEMP[7], CONST[3]
> 24: TEX TEMP[8], TEMP[1], SAMP[2], 2D
> 25: MUL TEMP[8], TEMP[8], CONST[4]
> 26: MUL TEMP[8], TEMP[8], CONST[5]
> 27: DP3 TEMP[3], -TEMP[6], TEMP[3]
> 28: ADD TEMP[3], TEMP[3], IMM[0].yyyy
> 29: POW TEMP[3], TEMP[3].xxxx, CONST[6].xxxx
> 30: RCP_SAT TEMP[9], TEMP[3].xxxx
> 31: TEX TEMP[3], TEMP[5], SAMP[3], RECT
> 32: MUL TEMP[3], TEMP[3], CONST[7]
> 33: TEX TEMP[2], TEMP[7], SAMP[4], CUBE
> 34: SUB TEMP[5], TEMP[2], TEMP[3]
> 35: MAD TEMP[5], TEMP[5], TEMP[9], TEMP[3]
> 36: ADD OUT[0], TEMP[5], TEMP[8]
> 37: MUL OUT[0].w, IN[1].wwww, TEMP[4].xxxx
> 38: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: RCP temp[10].w, input[0].wwww;
>  1: MUL temp[10].xyz, input[0], temp[10].wwww;
>  2: MAD temp[10].xyz, temp[10].xyz0, const[10].xyz0, const[11].xyz0;
>  3: MOV temp[0], temp[10];
>  4: SUB temp[0].y, const[8], temp[10];
>  5: TEX temp[1], input[2], 2D[0];
>  6: MOV temp[1].x, temp[1].wwww;
>  7: MAD temp[2], temp[1], const[9].xxxx, -const[9].yyyy;
>  8: MUL temp[2].xy, temp[2], const[1].zzzz;
>  9: MOV temp[2].z, const[9].zzzz;
> 10: DP3 temp[3].x, temp[2], temp[2];
> 11: ADD temp[3].x, const[9].yyyy, -temp[3].xxxx;
> 12: RSQ temp[3].x, temp[3].xxxx;
> 13: RCP temp[2].z, temp[3].xxxx;
> 14: TEX temp[4], input[2], 2D[1];
> 15: DP3 temp[3].x, temp[2], input[4];
> 16: DP3 temp[3].y, temp[2], input[5];
> 17: DP3 temp[3].z, temp[2], input[6];
> 18: MUL temp[5], const[1], const[2];
> 19: MAD temp[5], temp[5], temp[3], temp[0];
> 20: DP3 temp[6].w, input[3], input[3];
> 21: RSQ temp[6].w, temp[6].wwww;
> 22: MUL temp[6], input[3], temp[6].wwww;
> 23: DP3 temp[2], temp[6], temp[3];
> 24: MUL temp[2], temp[2], const[9].xxxx;
> 25: MAD temp[7], -temp[3], temp[2].xxxx, temp[6];
> 26: DP3_SAT temp[1], temp[7], const[3];
> 27: TEX temp[8], temp[1], 2D[2];
> 28: MUL temp[8], temp[8], const[4];
> 29: MUL temp[8], temp[8], const[5];
> 30: DP3 temp[3], -temp[6], temp[3];
> 31: ADD temp[3], temp[3], const[9].yyyy;
> 32: POW temp[3], temp[3].xxxx, const[6].xxxx;
> 33: RCP_SAT temp[9], temp[3].xxxx;
> 34: TEX temp[3], temp[5], RECT[3];
> 35: MUL temp[3], temp[3], const[7];
> 36: TEX temp[2], temp[7], CUBE[4];
> 37: SUB temp[5], temp[2], temp[3];
> 38: MAD temp[5], temp[5], temp[9], temp[3];
> 39: ADD output[0], temp[5], temp[8];
> 40: MUL output[0].w, input[1].wwww, temp[4].xxxx;
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: RCP temp[10].w, input[0].wwww;
>  1: MUL temp[10].xyz, input[0], temp[10].wwww;
>  2: MAD temp[10].xyz, temp[10].xyz0, const[10].xyz0, const[11].xyz0;
>  3: MOV temp[0], temp[10];
>  4: SUB temp[0].y, const[8], temp[10];
>  5: TEX temp[1], input[2], 2D[0];
>  6: MOV temp[1].x, temp[1].wwww;
>  7: MAD temp[2], temp[1], const[9].xxxx, -const[9].yyyy;
>  8: MUL temp[2].xy, temp[2], const[1].zzzz;
>  9: MOV temp[2].z, const[9].zzzz;
> 10: DP3 temp[3].x, temp[2], temp[2];
> 11: ADD temp[3].x, const[9].yyyy, -temp[3].xxxx;
> 12: RSQ temp[3].x, temp[3].xxxx;
> 13: RCP temp[2].z, temp[3].xxxx;
> 14: TEX temp[4], input[2], 2D[1];
> 15: DP3 temp[3].x, temp[2], input[4];
> 16: DP3 temp[3].y, temp[2], input[5];
> 17: DP3 temp[3].z, temp[2], input[6];
> 18: MUL temp[5], const[1], const[2];
> 19: MAD temp[5], temp[5], temp[3], temp[0];
> 20: DP3 temp[6].w, input[3], input[3];
> 21: RSQ temp[6].w, temp[6].wwww;
> 22: MUL temp[6], input[3], temp[6].wwww;
> 23: DP3 temp[2], temp[6], temp[3];
> 24: MUL temp[2], temp[2], const[9].xxxx;
> 25: MAD temp[7], -temp[3], temp[2].xxxx, temp[6];
> 26: DP3_SAT temp[1], temp[7], const[3];
> 27: TEX temp[8], temp[1], 2D[2];
> 28: MUL temp[8], temp[8], const[4];
> 29: MUL temp[8], temp[8], const[5];
> 30: DP3 temp[3], -temp[6], temp[3];
> 31: ADD temp[3], temp[3], const[9].yyyy;
> 32: POW temp[3], temp[3].xxxx, const[6].xxxx;
> 33: RCP_SAT temp[9], temp[3].xxxx;
> 34: TEX temp[3], temp[5], RECT[3];
> 35: MUL temp[3], temp[3], const[7];
> 36: TEX temp[2], temp[7], CUBE[4];
> 37: SUB temp[5], temp[2], temp[3];
> 38: MAD temp[5], temp[5], temp[9], temp[3];
> 39: ADD output[0], temp[5], temp[8];
> 40: MUL output[0].w, input[1].wwww, temp[4].xxxx;
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: RCP temp[10].w, input[0].wwww;
>  1: MUL temp[10].xyz, input[0], temp[10].wwww;
>  2: MAD temp[10].xyz, temp[10].xyz0, const[10].xyz0, const[11].xyz0;
>  3: MOV temp[0], temp[10];
>  4: SUB temp[0].y, const[8], temp[10];
>  5: TEX temp[1], input[2], 2D[0];
>  6: MOV temp[1].x, temp[1].wwww;
>  7: MAD temp[2], temp[1], const[9].xxxx, -const[9].yyyy;
>  8: MUL temp[2].xy, temp[2], const[1].zzzz;
>  9: MOV temp[2].z, const[9].zzzz;
> 10: DP3 temp[3].x, temp[2], temp[2];
> 11: ADD temp[3].x, const[9].yyyy, -temp[3].xxxx;
> 12: RSQ temp[3].x, temp[3].xxxx;
> 13: RCP temp[2].z, temp[3].xxxx;
> 14: TEX temp[4], input[2], 2D[1];
> 15: DP3 temp[3].x, temp[2], input[4];
> 16: DP3 temp[3].y, temp[2], input[5];
> 17: DP3 temp[3].z, temp[2], input[6];
> 18: MUL temp[5], const[1], const[2];
> 19: MAD temp[5], temp[5], temp[3], temp[0];
> 20: DP3 temp[6].w, input[3], input[3];
> 21: RSQ temp[6].w, temp[6].wwww;
> 22: MUL temp[6], input[3], temp[6].wwww;
> 23: DP3 temp[2], temp[6], temp[3];
> 24: MUL temp[2], temp[2], const[9].xxxx;
> 25: MAD temp[7], -temp[3], temp[2].xxxx, temp[6];
> 26: DP3_SAT temp[1], temp[7], const[3];
> 27: TEX temp[8], temp[1], 2D[2];
> 28: MUL temp[8], temp[8], const[4];
> 29: MUL temp[8], temp[8], const[5];
> 30: DP3 temp[3], -temp[6], temp[3];
> 31: ADD temp[3], temp[3], const[9].yyyy;
> 32: POW temp[3], temp[3].xxxx, const[6].xxxx;
> 33: RCP_SAT temp[9], temp[3].xxxx;
> 34: TEX temp[3], temp[5], RECT[3];
> 35: MUL temp[3], temp[3], const[7];
> 36: TEX temp[2], temp[7], CUBE[4];
> 37: SUB temp[5], temp[2], temp[3];
> 38: MAD temp[5], temp[5], temp[9], temp[3];
> 39: ADD output[0], temp[5], temp[8];
> 40: MUL output[0].w, input[1].wwww, temp[4].xxxx;
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: RCP temp[10].w, input[0].wwww;
>  1: MUL temp[10].xyz, input[0], temp[10].wwww;
>  2: MAD temp[10].xyz, temp[10].xyz0, const[10].xyz0, const[11].xyz0;
>  3: MOV temp[0], temp[10];
>  4: ADD temp[0].y, const[8], -temp[10];
>  5: TEX temp[1], input[2], 2D[0];
>  6: MOV temp[1].x, temp[1].wwww;
>  7: MAD temp[2], temp[1], const[9].xxxx, -const[9].yyyy;
>  8: MUL temp[2].xy, temp[2], const[1].zzzz;
>  9: MOV temp[2].z, const[9].zzzz;
> 10: DP3 temp[3].x, temp[2], temp[2];
> 11: ADD temp[3].x, const[9].yyyy, -temp[3].xxxx;
> 12: RSQ temp[3].x, |temp[3].xxxx|;
> 13: RCP temp[2].z, temp[3].xxxx;
> 14: TEX temp[4], input[2], 2D[1];
> 15: DP3 temp[3].x, temp[2], input[4];
> 16: DP3 temp[3].y, temp[2], input[5];
> 17: DP3 temp[3].z, temp[2], input[6];
> 18: MUL temp[5], const[1], const[2];
> 19: MAD temp[5], temp[5], temp[3], temp[0];
> 20: DP3 temp[6].w, input[3], input[3];
> 21: RSQ temp[6].w, |temp[6].wwww|;
> 22: MUL temp[6], input[3], temp[6].wwww;
> 23: DP3 temp[2], temp[6], temp[3];
> 24: MUL temp[2], temp[2], const[9].xxxx;
> 25: MAD temp[7], -temp[3], temp[2].xxxx, temp[6];
> 26: DP3_SAT temp[1], temp[7], const[3];
> 27: TEX temp[8], temp[1], 2D[2];
> 28: MUL temp[8], temp[8], const[4];
> 29: MUL temp[8], temp[8], const[5];
> 30: DP3 temp[3], -temp[6], temp[3];
> 31: ADD temp[3], temp[3], const[9].yyyy;
> 32: LG2 temp[11].w, temp[3].xxxx;
> 33: MUL temp[11].w, temp[11].wwww, const[6].xxxx;
> 34: EX2 temp[3], temp[11].wwww;
> 35: RCP_SAT temp[9], temp[3].xxxx;
> 36: TEX temp[3], temp[5], RECT[3];
> 37: MUL temp[3], temp[3], const[7];
> 38: TEX temp[2], temp[7], CUBE[4];
> 39: ADD temp[5], temp[2], -temp[3];
> 40: MAD temp[5], temp[5], temp[9], temp[3];
> 41: ADD output[0], temp[5], temp[8];
> 42: MUL output[0].w, input[1].wwww, temp[4].xxxx;
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: RCP temp[10].w, input[0].wwww;
>  1: MUL temp[10].xyz, input[0], temp[10].wwww;
>  2: MAD temp[10].xyz, temp[10].xyz0, const[10].xyz0, const[11].xyz0;
>  3: MOV temp[0], temp[10];
>  4: ADD temp[0].y, const[8], -temp[10];
>  5: TEX temp[1], input[2], 2D[0];
>  6: MOV temp[1].x, temp[1].wwww;
>  7: MAD temp[2], temp[1], const[9].xxxx, -const[9].yyyy;
>  8: MUL temp[2].xy, temp[2], const[1].zzzz;
>  9: MOV temp[2].z, const[9].zzzz;
> 10: DP3 temp[3].x, temp[2], temp[2];
> 11: ADD temp[3].x, const[9].yyyy, -temp[3].xxxx;
> 12: RSQ temp[3].x, |temp[3].xxxx|;
> 13: RCP temp[2].z, temp[3].xxxx;
> 14: TEX temp[4], input[2], 2D[1];
> 15: DP3 temp[3].x, temp[2], input[4];
> 16: DP3 temp[3].y, temp[2], input[5];
> 17: DP3 temp[3].z, temp[2], input[6];
> 18: MUL temp[5], const[1], const[2];
> 19: MAD temp[5], temp[5], temp[3], temp[0];
> 20: DP3 temp[6].w, input[3], input[3];
> 21: RSQ temp[6].w, |temp[6].wwww|;
> 22: MUL temp[6], input[3], temp[6].wwww;
> 23: DP3 temp[2], temp[6], temp[3];
> 24: MUL temp[2], temp[2], const[9].xxxx;
> 25: MAD temp[7], -temp[3], temp[2].xxxx, temp[6];
> 26: DP3_SAT temp[1], temp[7], const[3];
> 27: MUL temp[12], temp[1], const[12];
> 28: TEX temp[8], temp[12], 2D[2];
> 29: MUL temp[8], temp[8], const[4];
> 30: MUL temp[8], temp[8], const[5];
> 31: DP3 temp[3], -temp[6], temp[3];
> 32: ADD temp[3], temp[3], const[9].yyyy;
> 33: LG2 temp[11].w, temp[3].xxxx;
> 34: MUL temp[11].w, temp[11].wwww, const[6].xxxx;
> 35: EX2 temp[3], temp[11].wwww;
> 36: RCP_SAT temp[9], temp[3].xxxx;
> 37: MUL temp[13], temp[5], const[13];
> 38: TEX temp[3], temp[13], 2D[3];
> 39: MUL temp[3], temp[3], const[7];
> 40: TEX temp[2], temp[7], CUBE[4];
> 41: ADD temp[5], temp[2], -temp[3];
> 42: MAD temp[5], temp[5], temp[9], temp[3];
> 43: ADD output[0], temp[5], temp[8];
> 44: MUL output[0].w, input[1].wwww, temp[4].xxxx;
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: RCP temp[10].w, input[0].w___;
>  1: MUL temp[10].xy, input[0].xy__, temp[10].ww__;
>  2: MAD temp[10].xy, temp[10].xy__, const[10].xy__, const[11].xy__;
>  3: MOV temp[0].x, temp[10].x___;
>  4: ADD temp[0].y, const[8]._y__, -temp[10]._y__;
>  5: TEX temp[1].yw, input[2].xy__, 2D[0];
>  6: MOV temp[1].x, temp[1].w___;
>  7: MAD temp[2].xy, temp[1].xy__, const[9].xx__, -const[9].yy__;
>  8: MUL temp[2].xy, temp[2].xy__, const[1].zz__;
>  9: MOV temp[2].z, const[9].__z_;
> 10: DP3 temp[3].x, temp[2].xyz_, temp[2].xyz_;
> 11: ADD temp[3].x, const[9].y___, -temp[3].x___;
> 12: RSQ temp[3].x, |temp[3].x___|;
> 13: RCP temp[2].z, temp[3].x___;
> 14: TEX temp[4].x, input[2].xy__, 2D[1];
> 15: DP3 temp[3].x, temp[2].xyz_, input[4].xyz_;
> 16: DP3 temp[3].y, temp[2].xyz_, input[5].xyz_;
> 17: DP3 temp[3].z, temp[2].xyz_, input[6].xyz_;
> 18: MUL temp[5].xy, const[1].xy__, const[2].xy__;
> 19: MAD temp[5].xy, temp[5].xy__, temp[3].xy__, temp[0].xy__;
> 20: DP3 temp[6].w, input[3].xyz_, input[3].xyz_;
> 21: RSQ temp[6].w, |temp[6].w___|;
> 22: MUL temp[6].xyz, input[3].xyz_, temp[6].www_;
> 23: DP3 temp[2].x, temp[6].xyz_, temp[3].xyz_;
> 24: MUL temp[2].x, temp[2].x___, const[9].x___;
> 25: MAD temp[7].xyz, -temp[3].xyz_, temp[2].xxx_, temp[6].xyz_;
> 26: DP3_SAT temp[1].xy, temp[7].xyz_, const[3].xyz_;
> 27: MUL temp[12].xy, temp[1].xy__, const[12].xy__;
> 28: TEX temp[8].xyz, temp[12].xy__, 2D[2];
> 29: MUL temp[8].xyz, temp[8].xyz_, const[4].xyz_;
> 30: MUL temp[8].xyz, temp[8].xyz_, const[5].xyz_;
> 31: DP3 temp[3].x, -temp[6].xyz_, temp[3].xyz_;
> 32: ADD temp[3].x, temp[3].x___, const[9].y___;
> 33: LG2 temp[11].w, temp[3].x___;
> 34: MUL temp[11].w, temp[11].___w, const[6].___x;
> 35: EX2 temp[3].x, temp[11].w___;
> 36: RCP_SAT temp[9].xyz, temp[3].x___;
> 37: MUL temp[13].xy, temp[5].xy__, const[13].xy__;
> 38: TEX temp[3].xyz, temp[13].xy__, 2D[3];
> 39: MUL temp[3].xyz, temp[3].xyz_, const[7].xyz_;
> 40: TEX temp[2].xyz, temp[7].xyz_, CUBE[4];
> 41: ADD temp[5].xyz, temp[2].xyz_, -temp[3].xyz_;
> 42: MAD temp[5].xyz, temp[5].xyz_, temp[9].xyz_, temp[3].xyz_;
> 43: ADD output[0].xyz, temp[5].xyz_, temp[8].xyz_;
> 44: MUL output[0].w, input[1].___w, temp[4].___x;
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: RCP temp[10].w, input[0].w___;
>  1: MUL temp[10].xy, input[0].xy__, temp[10].ww__;
>  2: MAD temp[10].xy, temp[10].xy__, const[10].xy__, const[11].xy__;
>  3: MOV temp[0].x, temp[10].x___;
>  4: ADD temp[0].y, const[8]._y__, -temp[10]._y__;
>  5: TEX temp[1].yw, input[2].xy__, 2D[0];
>  6: MOV temp[1].x, temp[1].w___;
>  7: MAD temp[2].xy, temp[1].xy__, const[9].xx__, -none.11__;
>  8: MUL temp[2].xy, temp[2].xy__, const[1].zz__;
>  9: MOV temp[2].z, none.__0_;
> 10: DP3 temp[3].x, temp[2].xyz_, temp[2].xyz_;
> 11: ADD temp[3].x, none.1___, -temp[3].x___;
> 12: RSQ temp[3].x, |temp[3].x___|;
> 13: RCP temp[2].z, temp[3].x___;
> 14: TEX temp[4].x, input[2].xy__, 2D[1];
> 15: DP3 temp[3].x, temp[2].xyz_, input[4].xyz_;
> 16: DP3 temp[3].y, temp[2].xyz_, input[5].xyz_;
> 17: DP3 temp[3].z, temp[2].xyz_, input[6].xyz_;
> 18: MUL temp[5].xy, const[1].xy__, const[2].xy__;
> 19: MAD temp[5].xy, temp[5].xy__, temp[3].xy__, temp[0].xy__;
> 20: DP3 temp[6].w, input[3].xyz_, input[3].xyz_;
> 21: RSQ temp[6].w, |temp[6].w___|;
> 22: MUL temp[6].xyz, input[3].xyz_, temp[6].www_;
> 23: DP3 temp[2].x, temp[6].xyz_, temp[3].xyz_;
> 24: MUL temp[2].x, temp[2].x___, const[9].x___;
> 25: MAD temp[7].xyz, -temp[3].xyz_, temp[2].xxx_, temp[6].xyz_;
> 26: DP3_SAT temp[1].xy, temp[7].xyz_, const[3].xyz_;
> 27: MUL temp[12].xy, temp[1].xy__, const[12].xy__;
> 28: TEX temp[8].xyz, temp[12].xy__, 2D[2];
> 29: MUL temp[8].xyz, temp[8].xyz_, const[4].xyz_;
> 30: MUL temp[8].xyz, temp[8].xyz_, const[5].xyz_;
> 31: DP3 temp[3].x, -temp[6].xyz_, temp[3].xyz_;
> 32: ADD temp[3].x, temp[3].x___, none.1___;
> 33: LG2 temp[11].w, temp[3].x___;
> 34: MUL temp[11].w, temp[11].___w, const[6].___x;
> 35: EX2 temp[3].x, temp[11].w___;
> 36: RCP_SAT temp[9].xyz, temp[3].x___;
> 37: MUL temp[13].xy, temp[5].xy__, const[13].xy__;
> 38: TEX temp[3].xyz, temp[13].xy__, 2D[3];
> 39: MUL temp[3].xyz, temp[3].xyz_, const[7].xyz_;
> 40: TEX temp[2].xyz, temp[7].xyz_, CUBE[4];
> 41: ADD temp[5].xyz, temp[2].xyz_, -temp[3].xyz_;
> 42: MAD temp[5].xyz, temp[5].xyz_, temp[9].xyz_, temp[3].xyz_;
> 43: ADD output[0].xyz, temp[5].xyz_, temp[8].xyz_;
> 44: MUL output[0].w, input[1].___w, temp[4].___x;
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: RCP temp[10].w, input[0].w___;
>  1: MUL temp[10].xy, input[0].xy__, temp[10].ww__;
>  2: MAD temp[10].xy, temp[10].xy__, const[10].xy__, const[11].xy__;
>  3: MOV temp[0].x, temp[10].x___;
>  4: ADD temp[0].y, const[8]._y__, -temp[10]._y__;
>  5: TEX temp[1].yw, input[2].xy__, 2D[0];
>  6: MOV temp[1].x, temp[1].w___;
>  7: MAD temp[2].xy, temp[1].xy__, const[9].xx__, -none.11__;
>  8: MUL temp[2].xy, temp[2].xy__, const[1].zz__;
>  9: MOV temp[2].z, none.__0_;
> 10: DP3 temp[3].x, temp[2].xyz_, temp[2].xyz_;
> 11: ADD temp[3].x, none.1___, -temp[3].x___;
> 12: RSQ temp[3].x, |temp[3].x___|;
> 13: RCP temp[2].z, temp[3].x___;
> 14: TEX temp[4].x, input[2].xy__, 2D[1];
> 15: DP3 temp[3].x, temp[2].xyz_, input[4].xyz_;
> 16: DP3 temp[3].y, temp[2].xyz_, input[5].xyz_;
> 17: DP3 temp[3].z, temp[2].xyz_, input[6].xyz_;
> 18: MUL temp[5].xy, const[1].xy__, const[2].xy__;
> 19: MAD temp[5].xy, temp[5].xy__, temp[3].xy__, temp[0].xy__;
> 20: DP3 temp[6].w, input[3].xyz_, input[3].xyz_;
> 21: RSQ temp[6].w, |temp[6].w___|;
> 22: MUL temp[6].xyz, input[3].xyz_, temp[6].www_;
> 23: DP3 temp[2].x, temp[6].xyz_, temp[3].xyz_;
> 24: MUL temp[2].x, temp[2].x___, const[9].x___;
> 25: MAD temp[7].xyz, -temp[3].xyz_, temp[2].xxx_, temp[6].xyz_;
> 26: DP3_SAT temp[1].xy, temp[7].xyz_, const[3].xyz_;
> 27: MUL temp[12].xy, temp[1].xy__, const[12].xy__;
> 28: TEX temp[8].xyz, temp[12].xy__, 2D[2];
> 29: MUL temp[8].xyz, temp[8].xyz_, const[4].xyz_;
> 30: MUL temp[8].xyz, temp[8].xyz_, const[5].xyz_;
> 31: DP3 temp[3].x, -temp[6].xyz_, temp[3].xyz_;
> 32: ADD temp[3].x, temp[3].x___, none.1___;
> 33: LG2 temp[11].w, temp[3].x___;
> 34: MUL temp[11].w, temp[11].___w, const[6].___x;
> 35: EX2 temp[3].x, temp[11].w___;
> 36: RCP_SAT temp[9].xyz, temp[3].x___;
> 37: MUL temp[13].xy, temp[5].xy__, const[13].xy__;
> 38: TEX temp[3].xyz, temp[13].xy__, 2D[3];
> 39: MUL temp[3].xyz, temp[3].xyz_, const[7].xyz_;
> 40: TEX temp[2].xyz, temp[7].xyz_, CUBE[4];
> 41: ADD temp[5].xyz, temp[2].xyz_, -temp[3].xyz_;
> 42: MAD temp[5].xyz, temp[5].xyz_, temp[9].xyz_, temp[3].xyz_;
> 43: ADD output[0].xyz, temp[5].xyz_, temp[8].xyz_;
> 44: MUL output[0].w, input[1].___w, temp[4].___x;
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: src0.w = input[0]
>     RCP temp[10].w, src0.w
>  1: src0.xyz = input[0], src0.w = temp[10]
>     MAD temp[10].xy, src0.xy_, src0.ww_, src0.000
>  2: src0.xyz = temp[10], src1.xyz = const[10], src2.xyz = const[11]
>     MAD temp[10].xy, src0.xy_, src1.xy_, src2.xy_
>  3: src0.xyz = temp[10]
>     MAD temp[0].x, src0.x__, src0.111, src0.000
>  4: src0.xyz = const[8], src1.xyz = temp[10]
>     MAD temp[0].y, src0._y_, src0.111, -src1._y_
>  5: TEX temp[1].yw, input[2].xy__, 2D[0];
>  6: src0.w = temp[1]
>     MAD temp[1].x, src0.w__, src0.111, src0.000
>  7: src0.xyz = temp[1], src1.xyz = const[9]
>     MAD temp[2].xy, src0.xy_, src1.xx_, -src0.11_
>  8: src0.xyz = temp[2], src1.xyz = const[1]
>     MAD temp[2].xy, src0.xy_, src1.zz_, src0.000
>  9: 
>     MAD temp[2].z, src0.__0, src0.111, src0.000
> 10: src0.xyz = temp[2]
>     DP3 temp[3].x, src0.xyz, src0.xyz
> 11: src0.xyz = temp[3]
>     MAD temp[3].x, src0.1__, src0.111, -src0.x__
> 12: src0.xyz = temp[3]
>     REPL_ALPHA temp[3].x
>     RSQ, |src0.x|
> 13: src0.xyz = temp[3]
>     REPL_ALPHA temp[2].z
>     RCP, src0.x
> 14: TEX temp[4].x, input[2].xy__, 2D[1];
> 15: src0.xyz = temp[2], src1.xyz = input[4]
>     DP3 temp[3].x, src0.xyz, src1.xyz
> 16: src0.xyz = temp[2], src1.xyz = input[5]
>     DP3 temp[3].y, src0.xyz, src1.xyz
> 17: src0.xyz = temp[2], src1.xyz = input[6]
>     DP3 temp[3].z, src0.xyz, src1.xyz
> 18: src0.xyz = const[1], src1.xyz = const[2]
>     MAD temp[5].xy, src0.xy_, src1.xy_, src0.000
> 19: src0.xyz = temp[5], src1.xyz = temp[3], src2.xyz = temp[0]
>     MAD temp[5].xy, src0.xy_, src1.xy_, src2.xy_
> 20: src0.xyz = input[3]
>     DP3, src0.xyz, src0.xyz
>     DP3 temp[6].w, src0._, src0._
> 21: src0.w = temp[6]
>     RSQ temp[6].w, |src0.w|
> 22: src0.xyz = input[3], src0.w = temp[6]
>     MAD temp[6].xyz, src0.xyz, src0.www, src0.000
> 23: src0.xyz = temp[6], src1.xyz = temp[3]
>     DP3 temp[2].x, src0.xyz, src1.xyz
> 24: src0.xyz = temp[2], src1.xyz = const[9]
>     MAD temp[2].x, src0.x__, src1.x__, src0.000
> 25: src0.xyz = temp[3], src1.xyz = temp[2], src2.xyz = temp[6]
>     MAD temp[7].xyz, -src0.xyz, src1.xxx, src2.xyz
> 26: src0.xyz = temp[7], src1.xyz = const[3]
>     DP3_SAT temp[1].xy, src0.xyz, src1.xyz
> 27: src0.xyz = temp[1], src1.xyz = const[12]
>     MAD temp[12].xy, src0.xy_, src1.xy_, src0.000
> 28: TEX temp[8].xyz, temp[12].xy__, 2D[2];
> 29: src0.xyz = temp[8], src1.xyz = const[4]
>     MAD temp[8].xyz, src0.xyz, src1.xyz, src0.000
> 30: src0.xyz = temp[8], src1.xyz = const[5]
>     MAD temp[8].xyz, src0.xyz, src1.xyz, src0.000
> 31: src0.xyz = temp[6], src1.xyz = temp[3]
>     DP3 temp[3].x, -src0.xyz, src1.xyz
> 32: src0.xyz = temp[3]
>     MAD temp[3].x, src0.x__, src0.111, src0.1__
> 33: src0.xyz = temp[3]
>     LG2 temp[11].w, src0.x
> 34: src0.xyz = const[6], src0.w = temp[11]
>     MAD temp[11].w, src0.w, src0.x, src0.0
> 35: src0.w = temp[11]
>     REPL_ALPHA temp[3].x
>     EX2, src0.w
> 36: src0.xyz = temp[3]
>     REPL_ALPHA_SAT temp[9].xyz
>     RCP_SAT, src0.x
> 37: src0.xyz = temp[5], src1.xyz = const[13]
>     MAD temp[13].xy, src0.xy_, src1.xy_, src0.000
> 38: TEX temp[3].xyz, temp[13].xy__, 2D[3];
> 39: src0.xyz = temp[3], src1.xyz = const[7]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src0.000
> 40: TEX temp[2].xyz, temp[7].xyz_, CUBE[4];
> 41: src0.xyz = temp[2], src1.xyz = temp[3]
>     MAD temp[5].xyz, src0.xyz, src0.111, -src1.xyz
> 42: src0.xyz = temp[5], src1.xyz = temp[9], src2.xyz = temp[3]
>     MAD temp[5].xyz, src0.xyz, src1.xyz, src2.xyz
> 43: src0.xyz = temp[5], src1.xyz = temp[8]
>     MAD color[0].xyz, src0.xyz, src0.111, src1.xyz
> 44: src0.xyz = temp[4], src0.w = input[1]
>     MAD color[0].w, src0.w, src0.x, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[4].x, input[2].xy__, 2D[1];
>  2: TEX temp[1].yw, input[2].xy__, 2D[0];
>  3: src0.xyz = input[3]
>     DP3, src0.xyz, src0.xyz
>     DP3 temp[6].w, src0._, src0._
>  4: src0.w = temp[1], src1.w = temp[6]
>     MAD temp[1].x, src0.w__, src0.111, src0.000
>     RSQ temp[6].w, |src1.w|
>  5: src0.xyz = input[3], src0.w = temp[6], src1.xyz = temp[4], src1.w = input[1]
>     MAD temp[6].xyz, src0.xyz, src0.www, src0.000
>     MAD color[0].w, src1.w, src1.x, src0.0
>  6: src0.xyz = temp[1], src0.w = input[0], src1.xyz = const[9]
>     MAD temp[2].xy, src0.xy_, src1.xx_, -src0.11_
>     RCP temp[10].w, src0.w
>  7: src0.xyz = input[0], src0.w = temp[10]
>     MAD temp[10].xy, src0.xy_, src0.ww_, src0.000
>  8: src0.xyz = temp[10], src1.xyz = const[10], src2.xyz = const[11]
>     MAD temp[10].xy, src0.xy_, src1.xy_, src2.xy_
>  9: src0.xyz = const[8], src1.xyz = temp[10]
>     MAD temp[0].y, src0._y_, src0.111, -src1._y_
> 10: src0.xyz = temp[10]
>     MAD temp[0].x, src0.x__, src0.111, src0.000
> 11: src0.xyz = temp[2], src1.xyz = const[1]
>     MAD temp[2].xy, src0.xy_, src1.zz_, src0.000
> 12: src0.xyz = const[1], src1.xyz = const[2]
>     MAD temp[5].xy, src0.xy_, src1.xy_, src0.000
> 13: 
>     MAD temp[2].z, src0.__0, src0.111, src0.000
> 14: src0.xyz = temp[2]
>     DP3 temp[3].x, src0.xyz, src0.xyz
> 15: src0.xyz = temp[3]
>     MAD temp[3].x, src0.1__, src0.111, -src0.x__
> 16: src0.xyz = temp[3]
>     REPL_ALPHA temp[3].x
>     RSQ, |src0.x|
> 17: src0.xyz = temp[3]
>     REPL_ALPHA temp[2].z
>     RCP, src0.x
> 18: src0.xyz = temp[2], src1.xyz = input[4]
>     DP3 temp[3].x, src0.xyz, src1.xyz
> 19: src0.xyz = temp[2], src1.xyz = input[5]
>     DP3 temp[3].y, src0.xyz, src1.xyz
> 20: src0.xyz = temp[5], src1.xyz = temp[3], src2.xyz = temp[0]
>     MAD temp[5].xy, src0.xy_, src1.xy_, src2.xy_
> 21: src0.xyz = temp[5], src1.xyz = const[13]
>     MAD temp[13].xy, src0.xy_, src1.xy_, src0.000
> 22: src0.xyz = temp[2], src1.xyz = input[6]
>     DP3 temp[3].z, src0.xyz, src1.xyz
> 23: src0.xyz = temp[6], src1.xyz = temp[3]
>     DP3 temp[2].x, src0.xyz, src1.xyz
> 24: src0.xyz = temp[2], src1.xyz = const[9]
>     MAD temp[2].x, src0.x__, src1.x__, src0.000
> 25: src0.xyz = temp[3], src1.xyz = temp[2], src2.xyz = temp[6]
>     MAD temp[7].xyz, -src0.xyz, src1.xxx, src2.xyz
> 26: src0.xyz = temp[7], src1.xyz = const[3]
>     DP3_SAT temp[1].xy, src0.xyz, src1.xyz
> 27: src0.xyz = temp[1], src1.xyz = const[12]
>     MAD temp[12].xy, src0.xy_, src1.xy_, src0.000
> 28: src0.xyz = temp[6], src1.xyz = temp[3]
>     DP3 temp[3].x, -src0.xyz, src1.xyz
> 29: src0.xyz = temp[3]
>     MAD temp[3].x, src0.x__, src0.111, src0.1__
> 30: src0.xyz = temp[3]
>     LG2 temp[11].w, src0.x
> 31: src0.xyz = const[6], src0.w = temp[11]
>     MAD temp[11].w, src0.w, src0.x, src0.0
> 32: src0.w = temp[11]
>     REPL_ALPHA temp[3].x
>     EX2, src0.w
> 33: src0.xyz = temp[3]
>     REPL_ALPHA_SAT temp[9].xyz
>     RCP_SAT, src0.x
> 34: BEGIN_TEX;
> 35: TEX temp[3].xyz, temp[13].xy__, 2D[3];
> 36: TEX temp[8].xyz, temp[12].xy__, 2D[2];
> 37: TEX temp[2].xyz, temp[7].xyz_, CUBE[4];
> 38: src0.xyz = temp[8], src1.xyz = const[4]
>     MAD temp[8].xyz, src0.xyz, src1.xyz, src0.000
> 39: src0.xyz = temp[8], src1.xyz = const[5]
>     MAD temp[8].xyz, src0.xyz, src1.xyz, src0.000
> 40: src0.xyz = temp[3], src1.xyz = const[7]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src0.000
> 41: src0.xyz = temp[2], src1.xyz = temp[3]
>     MAD temp[5].xyz, src0.xyz, src0.111, -src1.xyz
> 42: src0.xyz = temp[5], src1.xyz = temp[9], src2.xyz = temp[3]
>     MAD temp[5].xyz, src0.xyz, src1.xyz, src2.xyz
> 43: src0.xyz = temp[5], src1.xyz = temp[8]
>     MAD color[0].xyz, src0.xyz, src0.111, src1.xyz
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[7].x, temp[1].xy__, 2D[1];
>  2: TEX temp[1].yw, temp[1].xy__, 2D[0];
>  3: src0.xyz = temp[2]
>     DP3, src0.xyz, src0.xyz
>     DP3 temp[8].w, src0._, src0._
>  4: src0.w = temp[1], src1.w = temp[8]
>     MAD temp[1].x, src0.w__, src0.111, src0.000
>     RSQ temp[8].w, |src1.w|
>  5: src0.xyz = temp[2], src0.w = temp[8], src1.xyz = temp[7], src1.w = temp[0]
>     MAD temp[8].xyz, src0.xyz, src0.www, src0.000
>     MAD color[0].w, src1.w, src1.x, src0.0
>  6: src0.xyz = temp[1], src0.w = temp[6], src1.xyz = const[9]
>     MAD temp[2].xy, src0.xy_, src1.xx_, -src0.11_
>     RCP temp[7].w, src0.w
>  7: src0.xyz = temp[6], src0.w = temp[7]
>     MAD temp[7].xy, src0.xy_, src0.ww_, src0.000
>  8: src0.xyz = temp[7], src1.xyz = const[10], src2.xyz = const[11]
>     MAD temp[7].xy, src0.xy_, src1.xy_, src2.xy_
>  9: src0.xyz = const[8], src1.xyz = temp[7]
>     MAD temp[0].y, src0._y_, src0.111, -src1._y_
> 10: src0.xyz = temp[7]
>     MAD temp[0].x, src0.x__, src0.111, src0.000
> 11: src0.xyz = temp[2], src1.xyz = const[1]
>     MAD temp[2].xy, src0.xy_, src1.zz_, src0.000
> 12: src0.xyz = const[1], src1.xyz = const[2]
>     MAD temp[7].xy, src0.xy_, src1.xy_, src0.000
> 13: 
>     MAD temp[2].z, src0.__0, src0.111, src0.000
> 14: src0.xyz = temp[2]
>     DP3 temp[6].x, src0.xyz, src0.xyz
> 15: src0.xyz = temp[6]
>     MAD temp[6].x, src0.1__, src0.111, -src0.x__
> 16: src0.xyz = temp[6]
>     REPL_ALPHA temp[6].x
>     RSQ, |src0.x|
> 17: src0.xyz = temp[6]
>     REPL_ALPHA temp[2].z
>     RCP, src0.x
> 18: src0.xyz = temp[2], src1.xyz = temp[3]
>     DP3 temp[6].x, src0.xyz, src1.xyz
> 19: src0.xyz = temp[2], src1.xyz = temp[4]
>     DP3 temp[6].y, src0.xyz, src1.xyz
> 20: src0.xyz = temp[7], src1.xyz = temp[6], src2.xyz = temp[0]
>     MAD temp[7].xy, src0.xy_, src1.xy_, src2.xy_
> 21: src0.xyz = temp[7], src1.xyz = const[13]
>     MAD temp[9].xy, src0.xy_, src1.xy_, src0.000
> 22: src0.xyz = temp[2], src1.xyz = temp[5]
>     DP3 temp[6].z, src0.xyz, src1.xyz
> 23: src0.xyz = temp[8], src1.xyz = temp[6]
>     DP3 temp[2].x, src0.xyz, src1.xyz
> 24: src0.xyz = temp[2], src1.xyz = const[9]
>     MAD temp[2].x, src0.x__, src1.x__, src0.000
> 25: src0.xyz = temp[6], src1.xyz = temp[2], src2.xyz = temp[8]
>     MAD temp[0].xyz, -src0.xyz, src1.xxx, src2.xyz
> 26: src0.xyz = temp[0], src1.xyz = const[3]
>     DP3_SAT temp[1].xy, src0.xyz, src1.xyz
> 27: src0.xyz = temp[1], src1.xyz = const[12]
>     MAD temp[4].xy, src0.xy_, src1.xy_, src0.000
> 28: src0.xyz = temp[8], src1.xyz = temp[6]
>     DP3 temp[6].x, -src0.xyz, src1.xyz
> 29: src0.xyz = temp[6]
>     MAD temp[6].x, src0.x__, src0.111, src0.1__
> 30: src0.xyz = temp[6]
>     LG2 temp[1].w, src0.x
> 31: src0.xyz = const[6], src0.w = temp[1]
>     MAD temp[1].w, src0.w, src0.x, src0.0
> 32: src0.w = temp[1]
>     REPL_ALPHA temp[6].x
>     EX2, src0.w
> 33: src0.xyz = temp[6]
>     REPL_ALPHA_SAT temp[3].xyz
>     RCP_SAT, src0.x
> 34: BEGIN_TEX;
> 35: TEX temp[6].xyz, temp[9].xy__, 2D[3];
> 36: TEX temp[1].xyz, temp[4].xy__, 2D[2];
> 37: TEX temp[2].xyz, temp[0].xyz_, CUBE[4];
> 38: src0.xyz = temp[1], src1.xyz = const[4]
>     MAD temp[1].xyz, src0.xyz, src1.xyz, src0.000
> 39: src0.xyz = temp[1], src1.xyz = const[5]
>     MAD temp[1].xyz, src0.xyz, src1.xyz, src0.000
> 40: src0.xyz = temp[6], src1.xyz = const[7]
>     MAD temp[6].xyz, src0.xyz, src1.xyz, src0.000
> 41: src0.xyz = temp[2], src1.xyz = temp[6]
>     MAD temp[7].xyz, src0.xyz, src0.111, -src1.xyz
> 42: src0.xyz = temp[7], src1.xyz = temp[3], src2.xyz = temp[6]
>     MAD temp[7].xyz, src0.xyz, src1.xyz, src2.xyz
> 43: src0.xyz = temp[7], src1.xyz = temp[1]
>     MAD color[0].xyz, src0.xyz, src0.111, src1.xyz
>pc=51*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 30, tex_end: 1  (code_addr: 00420780)
>  TEX:
>    TEX t7, t1, texture[1]   (000089c1)
>    TEX t1, t1, texture[0]   (00008041)
>  0: xyz:  t2  t0  t0 ->                      (00000002)
>       w:  t0  t0  t0 -> t8.w                 (00a00000)
>     xyz:   t2.xyz   t2.xyz   t2.xxx    op: 00804000
>       w:      1.0      1.0     t2.x    op: 00800891
>  1: xyz:  t0  t0  t0 -> t1.x                 (00840000)
>       w:  t1  t8  t0 -> t8.w                 (00a00201)
>     xyz:   t1.www      1.0      0.0    op: 00050a8c
>       w:   |t8.w|     t0.x     t0.x    op: 0580004a
>  2: xyz:  t2  t7  t0 -> t8.xyz               (03a001c2)
>       w:  t8  t0  t0 -> o0.w                 (01000008)
>     xyz:   t2.xyz   t8.www      0.0    op: 00050600
>       w:     t0.w     t7.x      0.0    op: 0004018a
>  3: xyz:  t1  c9  t0 -> t2.xy                (01880a41)
>       w:  t6  t0  t0 -> t7.w                 (009c0006)
>     xyz:   t1.xyz   c9.xxx     -1.0    op: 000d4280
>       w:     t6.w     t1.x     t1.x    op: 05000009
>  4: xyz:  t6  t0  t0 -> t7.xy                (019c0006)
>       w:  t7  t0  t0 ->                      (00000007)
>     xyz:   t6.xyz   t7.www      0.0    op: 00050600
>       w:     t6.x     t6.x     t6.x    op: 00000000
>  5: xyz:  t7 c10 c11 -> t7.xy                (019eba87)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t7.xyz  c10.xyz  c11.xyz    op: 00020200
>       w:     t7.x     t7.x     t7.x    op: 00000000
>  6: xyz:  c8  t7  t0 -> t0.y                 (010001e8)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   c8.xyz      1.0  -t7.xyz    op: 00090a80
>       w:     c8.x     c8.x     c8.x    op: 00000000
>  7: xyz:  t7  t0  t0 -> t0.x                 (00800007)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t7.xyz      1.0      0.0    op: 00050a80
>       w:     t7.x     t7.x     t7.x    op: 00000000
>  8: xyz:  t2  c1  t0 -> t2.xy                (01880842)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz   c1.zzz      0.0    op: 00050380
>       w:     t2.x     t2.x     t2.x    op: 00000000
>  9: xyz:  c1  c2  t0 -> t7.xy                (019c08a1)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   c1.xyz   c2.xyz      0.0    op: 00050200
>       w:     c1.x     c1.x     c1.x    op: 00000000
> 10: xyz:  t0  t0  t0 -> t2.z                 (02080000)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:      0.0      1.0      0.0    op: 00050a94
>       w:     t0.x     t0.x     t0.x    op: 00000000
> 11: xyz:  t2  t0  t0 -> t6.x                 (00980002)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz   t2.xyz   t2.xxx    op: 00804000
>       w:     t2.x     t2.x     t2.x    op: 00000000
> 12: xyz:  t6  t0  t0 -> t6.x                 (00980006)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:      1.0      1.0  -t6.xyz    op: 00080a95
>       w:     t6.x     t6.x     t6.x    op: 00000000
> 13: xyz:  t6  t0  t0 -> t6.x                 (00980006)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t6.xxx   t6.xxx   t6.xxx    op: 05004081
>       w:   |t6.x|     t6.x     t6.x    op: 05800040
> 14: xyz:  t6  t0  t0 -> t2.z                 (02080006)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t6.xxx   t6.xxx   t6.xxx    op: 05004081
>       w:     t6.x     t6.x     t6.x    op: 05000000
> 15: xyz:  t2  t3  t0 -> t6.x                 (009800c2)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz   t3.xyz   t2.xxx    op: 00804200
>       w:     t2.x     t2.x     t2.x    op: 00000000
> 16: xyz:  t2  t4  t0 -> t6.y                 (01180102)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz   t4.xyz   t2.xxx    op: 00804200
>       w:     t2.x     t2.x     t2.x    op: 00000000
> 17: xyz:  t7  t6  t0 -> t7.xy                (019c0187)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t7.xyz   t6.xyz   t0.xyz    op: 00020200
>       w:     t7.x     t7.x     t7.x    op: 00000000
> 18: xyz:  t7 c13  t0 -> t9.xy                (01a40b47)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t7.xyz  c13.xyz      0.0    op: 00050200
>       w:     t7.x     t7.x     t7.x    op: 00000000
> 19: xyz:  t2  t5  t0 -> t6.z                 (02180142)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz   t5.xyz   t2.xxx    op: 00804200
>       w:     t2.x     t2.x     t2.x    op: 00000000
> 20: xyz:  t8  t6  t0 -> t2.x                 (00880188)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t8.xyz   t6.xyz   t8.xxx    op: 00804200
>       w:     t8.x     t8.x     t8.x    op: 00000000
> 21: xyz:  t2  c9  t0 -> t2.x                 (00880a42)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz   c9.xyz      0.0    op: 00050200
>       w:     t2.x     t2.x     t2.x    op: 00000000
> 22: xyz:  t6  t2  t8 -> t0.xyz               (03808086)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:  -t6.xyz   t2.xxx   t8.xyz    op: 000202a0
>       w:     t6.x     t6.x     t6.x    op: 00000000
> 23: xyz:  t0  c3  t0 -> t1.xy                (018408c0)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xyz   c3.xyz   t0.xxx    op: 40804200
>       w:     t0.x     t0.x     t0.x    op: 00000000
> 24: xyz:  t1 c12  t0 -> t4.xy                (01900b01)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz  c12.xyz      0.0    op: 00050200
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 25: xyz:  t8  t6  t0 -> t6.x                 (00980188)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:  -t8.xyz   t6.xyz   t8.xxx    op: 00804220
>       w:     t8.x     t8.x     t8.x    op: 00000000
> 26: xyz:  t6  t0  t0 -> t6.x                 (00980006)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t6.xyz      1.0      1.0    op: 00054a80
>       w:     t6.x     t6.x     t6.x    op: 00000000
> 27: xyz:  t6  t0  t0 ->                      (00000006)
>       w:  t0  t0  t0 -> t1.w                 (00840000)
>     xyz:   t6.xxx   t6.xxx   t6.xxx    op: 00004081
>       w:     t6.x     t6.x     t6.x    op: 04800000
> 28: xyz:  c6  t0  t0 ->                      (00000026)
>       w:  t1  t0  t0 -> t1.w                 (00840001)
>     xyz:   c6.xxx   c6.xxx   c6.xxx    op: 00004081
>       w:     t1.w     c6.x      0.0    op: 00040009
> 29: xyz:  t0  t0  t0 -> t6.x                 (00980000)
>       w:  t1  t0  t0 ->                      (00000001)
>     xyz:   t0.xxx   t0.xxx   t0.xxx    op: 05004081
>       w:     t1.w     t0.x     t0.x    op: 04000009
> 30: xyz:  t6  t0  t0 -> t3.xyz               (038c0006)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t6.xxx   t6.xxx   t6.xxx    op: 45004081
>       w:     t6.x     t6.x     t6.x    op: 45000000
>NODE 1: alu_offset: 31, tex_offset: 2, alu_end: 5, tex_end: 2  (code_addr: 0044215f)
>  TEX:
>    TEX t6, t9, texture[3]   (00009989)
>    TEX t1, t4, texture[2]   (00009044)
>    TEX t2, t0, texture[4]   (0000a080)
> 31: xyz:  t1  c4  t0 -> t1.xyz               (03840901)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   c4.xyz      0.0    op: 00050200
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 32: xyz:  t1  c5  t0 -> t1.xyz               (03840941)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   c5.xyz      0.0    op: 00050200
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 33: xyz:  t6  c7  t0 -> t6.xyz               (039809c6)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t6.xyz   c7.xyz      0.0    op: 00050200
>       w:     t6.x     t6.x     t6.x    op: 00000000
> 34: xyz:  t2  t6  t0 -> t7.xyz               (039c0182)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz      1.0  -t6.xyz    op: 00090a80
>       w:     t2.x     t2.x     t2.x    op: 00000000
> 35: xyz:  t7  t3  t6 -> t7.xyz               (039c60c7)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t7.xyz   t3.xyz   t6.xyz    op: 00020200
>       w:     t7.x     t7.x     t7.x    op: 00000000
> 36: xyz:  t7  t1  t0 -> o0.xyz               (1c000047)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t7.xyz      1.0   t1.xyz    op: 00010a80
>       w:     t7.x     t7.x     t7.x    op: 00000000
>r300: Initial fragment program
>FRAG
>DCL IN[0], POSITION, LINEAR
>DCL IN[1], COLOR, LINEAR
>DCL IN[2], GENERIC[0], PERSPECTIVE
>DCL IN[3], GENERIC[2], PERSPECTIVE
>DCL IN[4], GENERIC[3], PERSPECTIVE
>DCL IN[5], GENERIC[4], PERSPECTIVE
>DCL IN[6], GENERIC[5], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL SAMP[1]
>DCL SAMP[2]
>DCL SAMP[3]
>DCL SAMP[4]
>DCL CONST[7..8]
>DCL CONST[1..6]
>DCL TEMP[0..9]
>IMM FLT32 {    2.0000,     1.0000,     0.0000,     0.0000}
>  0: MOV TEMP[0], IN[0]
>  1: SUB TEMP[0].y, CONST[8], IN[0]
>  2: TEX TEMP[1], IN[2], SAMP[0], 2D
>  3: MOV TEMP[1].x, TEMP[1].wwww
>  4: MAD TEMP[2], TEMP[1], IMM[0].xxxx, -IMM[0].yyyy
>  5: MUL TEMP[2].xy, TEMP[2], CONST[1].zzzz
>  6: MOV TEMP[2].z, IMM[0].zzzz
>  7: DP3 TEMP[3].x, TEMP[2], TEMP[2]
>  8: ADD TEMP[3].x, IMM[0].yyyy, -TEMP[3].xxxx
>  9: RSQ TEMP[3].x, TEMP[3].xxxx
> 10: RCP TEMP[2].z, TEMP[3].xxxx
> 11: TEX TEMP[4], IN[2], SAMP[1], 2D
> 12: DP3 TEMP[3].x, TEMP[2], IN[4]
> 13: DP3 TEMP[3].y, TEMP[2], IN[5]
> 14: DP3 TEMP[3].z, TEMP[2], IN[6]
> 15: MUL TEMP[5], CONST[1], CONST[2]
> 16: MAD TEMP[5], TEMP[5], TEMP[3], TEMP[0]
> 17: DP3 TEMP[6].w, IN[3], IN[3]
> 18: RSQ TEMP[6].w, TEMP[6].wwww
> 19: MUL TEMP[6], IN[3], TEMP[6].wwww
> 20: DP3 TEMP[2], TEMP[6], TEMP[3]
> 21: MUL TEMP[2], TEMP[2], IMM[0].xxxx
> 22: MAD TEMP[7], -TEMP[3], TEMP[2].xxxx, TEMP[6]
> 23: DP3_SAT TEMP[1], TEMP[7], CONST[3]
> 24: TEX TEMP[8], TEMP[1], SAMP[2], 2D
> 25: MUL TEMP[8], TEMP[8], CONST[4]
> 26: MUL TEMP[8], TEMP[8], CONST[5]
> 27: DP3 TEMP[3], -TEMP[6], TEMP[3]
> 28: ADD TEMP[3], TEMP[3], IMM[0].yyyy
> 29: POW TEMP[3], TEMP[3].xxxx, CONST[6].xxxx
> 30: RCP_SAT TEMP[9], TEMP[3].xxxx
> 31: TEX TEMP[3], TEMP[5], SAMP[3], RECT
> 32: MUL TEMP[3], TEMP[3], CONST[7]
> 33: TEX TEMP[2], TEMP[7], SAMP[4], CUBE
> 34: SUB TEMP[5], TEMP[2], TEMP[3]
> 35: MAD TEMP[5], TEMP[5], TEMP[9], TEMP[3]
> 36: ADD OUT[0], TEMP[5], TEMP[8]
> 37: MUL OUT[0].w, IN[1].wwww, TEMP[4].xxxx
> 38: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: RCP temp[10].w, input[0].wwww;
>  1: MUL temp[10].xyz, input[0], temp[10].wwww;
>  2: MAD temp[10].xyz, temp[10].xyz0, const[10].xyz0, const[11].xyz0;
>  3: MOV temp[0], temp[10];
>  4: SUB temp[0].y, const[8], temp[10];
>  5: TEX temp[1], input[2], 2D[0];
>  6: MOV temp[1].x, temp[1].wwww;
>  7: MAD temp[2], temp[1], const[9].xxxx, -const[9].yyyy;
>  8: MUL temp[2].xy, temp[2], const[1].zzzz;
>  9: MOV temp[2].z, const[9].zzzz;
> 10: DP3 temp[3].x, temp[2], temp[2];
> 11: ADD temp[3].x, const[9].yyyy, -temp[3].xxxx;
> 12: RSQ temp[3].x, temp[3].xxxx;
> 13: RCP temp[2].z, temp[3].xxxx;
> 14: TEX temp[4], input[2], 2D[1];
> 15: DP3 temp[3].x, temp[2], input[4];
> 16: DP3 temp[3].y, temp[2], input[5];
> 17: DP3 temp[3].z, temp[2], input[6];
> 18: MUL temp[5], const[1], const[2];
> 19: MAD temp[5], temp[5], temp[3], temp[0];
> 20: DP3 temp[6].w, input[3], input[3];
> 21: RSQ temp[6].w, temp[6].wwww;
> 22: MUL temp[6], input[3], temp[6].wwww;
> 23: DP3 temp[2], temp[6], temp[3];
> 24: MUL temp[2], temp[2], const[9].xxxx;
> 25: MAD temp[7], -temp[3], temp[2].xxxx, temp[6];
> 26: DP3_SAT temp[1], temp[7], const[3];
> 27: TEX temp[8], temp[1], 2D[2];
> 28: MUL temp[8], temp[8], const[4];
> 29: MUL temp[8], temp[8], const[5];
> 30: DP3 temp[3], -temp[6], temp[3];
> 31: ADD temp[3], temp[3], const[9].yyyy;
> 32: POW temp[3], temp[3].xxxx, const[6].xxxx;
> 33: RCP_SAT temp[9], temp[3].xxxx;
> 34: TEX temp[3], temp[5], RECT[3];
> 35: MUL temp[3], temp[3], const[7];
> 36: TEX temp[2], temp[7], CUBE[4];
> 37: SUB temp[5], temp[2], temp[3];
> 38: MAD temp[5], temp[5], temp[9], temp[3];
> 39: ADD output[0], temp[5], temp[8];
> 40: MUL output[0].w, input[1].wwww, temp[4].xxxx;
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: RCP temp[10].w, input[0].wwww;
>  1: MUL temp[10].xyz, input[0], temp[10].wwww;
>  2: MAD temp[10].xyz, temp[10].xyz0, const[10].xyz0, const[11].xyz0;
>  3: MOV temp[0], temp[10];
>  4: SUB temp[0].y, const[8], temp[10];
>  5: TEX temp[1], input[2], 2D[0];
>  6: MOV temp[1].x, temp[1].wwww;
>  7: MAD temp[2], temp[1], const[9].xxxx, -const[9].yyyy;
>  8: MUL temp[2].xy, temp[2], const[1].zzzz;
>  9: MOV temp[2].z, const[9].zzzz;
> 10: DP3 temp[3].x, temp[2], temp[2];
> 11: ADD temp[3].x, const[9].yyyy, -temp[3].xxxx;
> 12: RSQ temp[3].x, temp[3].xxxx;
> 13: RCP temp[2].z, temp[3].xxxx;
> 14: TEX temp[4], input[2], 2D[1];
> 15: DP3 temp[3].x, temp[2], input[4];
> 16: DP3 temp[3].y, temp[2], input[5];
> 17: DP3 temp[3].z, temp[2], input[6];
> 18: MUL temp[5], const[1], const[2];
> 19: MAD temp[5], temp[5], temp[3], temp[0];
> 20: DP3 temp[6].w, input[3], input[3];
> 21: RSQ temp[6].w, temp[6].wwww;
> 22: MUL temp[6], input[3], temp[6].wwww;
> 23: DP3 temp[2], temp[6], temp[3];
> 24: MUL temp[2], temp[2], const[9].xxxx;
> 25: MAD temp[7], -temp[3], temp[2].xxxx, temp[6];
> 26: DP3_SAT temp[1], temp[7], const[3];
> 27: TEX temp[8], temp[1], 2D[2];
> 28: MUL temp[8], temp[8], const[4];
> 29: MUL temp[8], temp[8], const[5];
> 30: DP3 temp[3], -temp[6], temp[3];
> 31: ADD temp[3], temp[3], const[9].yyyy;
> 32: POW temp[3], temp[3].xxxx, const[6].xxxx;
> 33: RCP_SAT temp[9], temp[3].xxxx;
> 34: TEX temp[3], temp[5], RECT[3];
> 35: MUL temp[3], temp[3], const[7];
> 36: TEX temp[2], temp[7], CUBE[4];
> 37: SUB temp[5], temp[2], temp[3];
> 38: MAD temp[5], temp[5], temp[9], temp[3];
> 39: ADD output[0], temp[5], temp[8];
> 40: MUL output[0].w, input[1].wwww, temp[4].xxxx;
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: RCP temp[10].w, input[0].wwww;
>  1: MUL temp[10].xyz, input[0], temp[10].wwww;
>  2: MAD temp[10].xyz, temp[10].xyz0, const[10].xyz0, const[11].xyz0;
>  3: MOV temp[0], temp[10];
>  4: SUB temp[0].y, const[8], temp[10];
>  5: TEX temp[1], input[2], 2D[0];
>  6: MOV temp[1].x, temp[1].wwww;
>  7: MAD temp[2], temp[1], const[9].xxxx, -const[9].yyyy;
>  8: MUL temp[2].xy, temp[2], const[1].zzzz;
>  9: MOV temp[2].z, const[9].zzzz;
> 10: DP3 temp[3].x, temp[2], temp[2];
> 11: ADD temp[3].x, const[9].yyyy, -temp[3].xxxx;
> 12: RSQ temp[3].x, temp[3].xxxx;
> 13: RCP temp[2].z, temp[3].xxxx;
> 14: TEX temp[4], input[2], 2D[1];
> 15: DP3 temp[3].x, temp[2], input[4];
> 16: DP3 temp[3].y, temp[2], input[5];
> 17: DP3 temp[3].z, temp[2], input[6];
> 18: MUL temp[5], const[1], const[2];
> 19: MAD temp[5], temp[5], temp[3], temp[0];
> 20: DP3 temp[6].w, input[3], input[3];
> 21: RSQ temp[6].w, temp[6].wwww;
> 22: MUL temp[6], input[3], temp[6].wwww;
> 23: DP3 temp[2], temp[6], temp[3];
> 24: MUL temp[2], temp[2], const[9].xxxx;
> 25: MAD temp[7], -temp[3], temp[2].xxxx, temp[6];
> 26: DP3_SAT temp[1], temp[7], const[3];
> 27: TEX temp[8], temp[1], 2D[2];
> 28: MUL temp[8], temp[8], const[4];
> 29: MUL temp[8], temp[8], const[5];
> 30: DP3 temp[3], -temp[6], temp[3];
> 31: ADD temp[3], temp[3], const[9].yyyy;
> 32: POW temp[3], temp[3].xxxx, const[6].xxxx;
> 33: RCP_SAT temp[9], temp[3].xxxx;
> 34: TEX temp[3], temp[5], RECT[3];
> 35: MUL temp[3], temp[3], const[7];
> 36: TEX temp[2], temp[7], CUBE[4];
> 37: SUB temp[5], temp[2], temp[3];
> 38: MAD temp[5], temp[5], temp[9], temp[3];
> 39: ADD output[0], temp[5], temp[8];
> 40: MUL output[0].w, input[1].wwww, temp[4].xxxx;
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: RCP temp[10].w, input[0].wwww;
>  1: MUL temp[10].xyz, input[0], temp[10].wwww;
>  2: MAD temp[10].xyz, temp[10].xyz0, const[10].xyz0, const[11].xyz0;
>  3: MOV temp[0], temp[10];
>  4: ADD temp[0].y, const[8], -temp[10];
>  5: TEX temp[1], input[2], 2D[0];
>  6: MOV temp[1].x, temp[1].wwww;
>  7: MAD temp[2], temp[1], const[9].xxxx, -const[9].yyyy;
>  8: MUL temp[2].xy, temp[2], const[1].zzzz;
>  9: MOV temp[2].z, const[9].zzzz;
> 10: DP3 temp[3].x, temp[2], temp[2];
> 11: ADD temp[3].x, const[9].yyyy, -temp[3].xxxx;
> 12: RSQ temp[3].x, |temp[3].xxxx|;
> 13: RCP temp[2].z, temp[3].xxxx;
> 14: TEX temp[4], input[2], 2D[1];
> 15: DP3 temp[3].x, temp[2], input[4];
> 16: DP3 temp[3].y, temp[2], input[5];
> 17: DP3 temp[3].z, temp[2], input[6];
> 18: MUL temp[5], const[1], const[2];
> 19: MAD temp[5], temp[5], temp[3], temp[0];
> 20: DP3 temp[6].w, input[3], input[3];
> 21: RSQ temp[6].w, |temp[6].wwww|;
> 22: MUL temp[6], input[3], temp[6].wwww;
> 23: DP3 temp[2], temp[6], temp[3];
> 24: MUL temp[2], temp[2], const[9].xxxx;
> 25: MAD temp[7], -temp[3], temp[2].xxxx, temp[6];
> 26: DP3_SAT temp[1], temp[7], const[3];
> 27: TEX temp[8], temp[1], 2D[2];
> 28: MUL temp[8], temp[8], const[4];
> 29: MUL temp[8], temp[8], const[5];
> 30: DP3 temp[3], -temp[6], temp[3];
> 31: ADD temp[3], temp[3], const[9].yyyy;
> 32: LG2 temp[11].w, temp[3].xxxx;
> 33: MUL temp[11].w, temp[11].wwww, const[6].xxxx;
> 34: EX2 temp[3], temp[11].wwww;
> 35: RCP_SAT temp[9], temp[3].xxxx;
> 36: TEX temp[3], temp[5], RECT[3];
> 37: MUL temp[3], temp[3], const[7];
> 38: TEX temp[2], temp[7], CUBE[4];
> 39: ADD temp[5], temp[2], -temp[3];
> 40: MAD temp[5], temp[5], temp[9], temp[3];
> 41: ADD output[0], temp[5], temp[8];
> 42: MUL output[0].w, input[1].wwww, temp[4].xxxx;
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: RCP temp[10].w, input[0].wwww;
>  1: MUL temp[10].xyz, input[0], temp[10].wwww;
>  2: MAD temp[10].xyz, temp[10].xyz0, const[10].xyz0, const[11].xyz0;
>  3: MOV temp[0], temp[10];
>  4: ADD temp[0].y, const[8], -temp[10];
>  5: TEX temp[1], input[2], 2D[0];
>  6: MOV temp[1].x, temp[1].wwww;
>  7: MAD temp[2], temp[1], const[9].xxxx, -const[9].yyyy;
>  8: MUL temp[2].xy, temp[2], const[1].zzzz;
>  9: MOV temp[2].z, const[9].zzzz;
> 10: DP3 temp[3].x, temp[2], temp[2];
> 11: ADD temp[3].x, const[9].yyyy, -temp[3].xxxx;
> 12: RSQ temp[3].x, |temp[3].xxxx|;
> 13: RCP temp[2].z, temp[3].xxxx;
> 14: TEX temp[4], input[2], 2D[1];
> 15: DP3 temp[3].x, temp[2], input[4];
> 16: DP3 temp[3].y, temp[2], input[5];
> 17: DP3 temp[3].z, temp[2], input[6];
> 18: MUL temp[5], const[1], const[2];
> 19: MAD temp[5], temp[5], temp[3], temp[0];
> 20: DP3 temp[6].w, input[3], input[3];
> 21: RSQ temp[6].w, |temp[6].wwww|;
> 22: MUL temp[6], input[3], temp[6].wwww;
> 23: DP3 temp[2], temp[6], temp[3];
> 24: MUL temp[2], temp[2], const[9].xxxx;
> 25: MAD temp[7], -temp[3], temp[2].xxxx, temp[6];
> 26: DP3_SAT temp[1], temp[7], const[3];
> 27: TEX temp[8], temp[1], 2D[2];
> 28: MUL temp[8], temp[8], const[4];
> 29: MUL temp[8], temp[8], const[5];
> 30: DP3 temp[3], -temp[6], temp[3];
> 31: ADD temp[3], temp[3], const[9].yyyy;
> 32: LG2 temp[11].w, temp[3].xxxx;
> 33: MUL temp[11].w, temp[11].wwww, const[6].xxxx;
> 34: EX2 temp[3], temp[11].wwww;
> 35: RCP_SAT temp[9], temp[3].xxxx;
> 36: MUL temp[12], temp[5], const[12];
> 37: TEX temp[3], temp[12], 2D[3];
> 38: MUL temp[3], temp[3], const[7];
> 39: TEX temp[2], temp[7], CUBE[4];
> 40: ADD temp[5], temp[2], -temp[3];
> 41: MAD temp[5], temp[5], temp[9], temp[3];
> 42: ADD output[0], temp[5], temp[8];
> 43: MUL output[0].w, input[1].wwww, temp[4].xxxx;
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: RCP temp[10].w, input[0].w___;
>  1: MUL temp[10].xy, input[0].xy__, temp[10].ww__;
>  2: MAD temp[10].xy, temp[10].xy__, const[10].xy__, const[11].xy__;
>  3: MOV temp[0].x, temp[10].x___;
>  4: ADD temp[0].y, const[8]._y__, -temp[10]._y__;
>  5: TEX temp[1].yw, input[2].xy__, 2D[0];
>  6: MOV temp[1].x, temp[1].w___;
>  7: MAD temp[2].xy, temp[1].xy__, const[9].xx__, -const[9].yy__;
>  8: MUL temp[2].xy, temp[2].xy__, const[1].zz__;
>  9: MOV temp[2].z, const[9].__z_;
> 10: DP3 temp[3].x, temp[2].xyz_, temp[2].xyz_;
> 11: ADD temp[3].x, const[9].y___, -temp[3].x___;
> 12: RSQ temp[3].x, |temp[3].x___|;
> 13: RCP temp[2].z, temp[3].x___;
> 14: TEX temp[4].x, input[2].xy__, 2D[1];
> 15: DP3 temp[3].x, temp[2].xyz_, input[4].xyz_;
> 16: DP3 temp[3].y, temp[2].xyz_, input[5].xyz_;
> 17: DP3 temp[3].z, temp[2].xyz_, input[6].xyz_;
> 18: MUL temp[5].xy, const[1].xy__, const[2].xy__;
> 19: MAD temp[5].xy, temp[5].xy__, temp[3].xy__, temp[0].xy__;
> 20: DP3 temp[6].w, input[3].xyz_, input[3].xyz_;
> 21: RSQ temp[6].w, |temp[6].w___|;
> 22: MUL temp[6].xyz, input[3].xyz_, temp[6].www_;
> 23: DP3 temp[2].x, temp[6].xyz_, temp[3].xyz_;
> 24: MUL temp[2].x, temp[2].x___, const[9].x___;
> 25: MAD temp[7].xyz, -temp[3].xyz_, temp[2].xxx_, temp[6].xyz_;
> 26: DP3_SAT temp[1].xy, temp[7].xyz_, const[3].xyz_;
> 27: TEX temp[8].xyz, temp[1].xy__, 2D[2];
> 28: MUL temp[8].xyz, temp[8].xyz_, const[4].xyz_;
> 29: MUL temp[8].xyz, temp[8].xyz_, const[5].xyz_;
> 30: DP3 temp[3].x, -temp[6].xyz_, temp[3].xyz_;
> 31: ADD temp[3].x, temp[3].x___, const[9].y___;
> 32: LG2 temp[11].w, temp[3].x___;
> 33: MUL temp[11].w, temp[11].___w, const[6].___x;
> 34: EX2 temp[3].x, temp[11].w___;
> 35: RCP_SAT temp[9].xyz, temp[3].x___;
> 36: MUL temp[12].xy, temp[5].xy__, const[12].xy__;
> 37: TEX temp[3].xyz, temp[12].xy__, 2D[3];
> 38: MUL temp[3].xyz, temp[3].xyz_, const[7].xyz_;
> 39: TEX temp[2].xyz, temp[7].xyz_, CUBE[4];
> 40: ADD temp[5].xyz, temp[2].xyz_, -temp[3].xyz_;
> 41: MAD temp[5].xyz, temp[5].xyz_, temp[9].xyz_, temp[3].xyz_;
> 42: ADD output[0].xyz, temp[5].xyz_, temp[8].xyz_;
> 43: MUL output[0].w, input[1].___w, temp[4].___x;
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: RCP temp[10].w, input[0].w___;
>  1: MUL temp[10].xy, input[0].xy__, temp[10].ww__;
>  2: MAD temp[10].xy, temp[10].xy__, const[10].xy__, const[11].xy__;
>  3: MOV temp[0].x, temp[10].x___;
>  4: ADD temp[0].y, const[8]._y__, -temp[10]._y__;
>  5: TEX temp[1].yw, input[2].xy__, 2D[0];
>  6: MOV temp[1].x, temp[1].w___;
>  7: MAD temp[2].xy, temp[1].xy__, const[9].xx__, -none.11__;
>  8: MUL temp[2].xy, temp[2].xy__, const[1].zz__;
>  9: MOV temp[2].z, none.__0_;
> 10: DP3 temp[3].x, temp[2].xyz_, temp[2].xyz_;
> 11: ADD temp[3].x, none.1___, -temp[3].x___;
> 12: RSQ temp[3].x, |temp[3].x___|;
> 13: RCP temp[2].z, temp[3].x___;
> 14: TEX temp[4].x, input[2].xy__, 2D[1];
> 15: DP3 temp[3].x, temp[2].xyz_, input[4].xyz_;
> 16: DP3 temp[3].y, temp[2].xyz_, input[5].xyz_;
> 17: DP3 temp[3].z, temp[2].xyz_, input[6].xyz_;
> 18: MUL temp[5].xy, const[1].xy__, const[2].xy__;
> 19: MAD temp[5].xy, temp[5].xy__, temp[3].xy__, temp[0].xy__;
> 20: DP3 temp[6].w, input[3].xyz_, input[3].xyz_;
> 21: RSQ temp[6].w, |temp[6].w___|;
> 22: MUL temp[6].xyz, input[3].xyz_, temp[6].www_;
> 23: DP3 temp[2].x, temp[6].xyz_, temp[3].xyz_;
> 24: MUL temp[2].x, temp[2].x___, const[9].x___;
> 25: MAD temp[7].xyz, -temp[3].xyz_, temp[2].xxx_, temp[6].xyz_;
> 26: DP3_SAT temp[1].xy, temp[7].xyz_, const[3].xyz_;
> 27: TEX temp[8].xyz, temp[1].xy__, 2D[2];
> 28: MUL temp[8].xyz, temp[8].xyz_, const[4].xyz_;
> 29: MUL temp[8].xyz, temp[8].xyz_, const[5].xyz_;
> 30: DP3 temp[3].x, -temp[6].xyz_, temp[3].xyz_;
> 31: ADD temp[3].x, temp[3].x___, none.1___;
> 32: LG2 temp[11].w, temp[3].x___;
> 33: MUL temp[11].w, temp[11].___w, const[6].___x;
> 34: EX2 temp[3].x, temp[11].w___;
> 35: RCP_SAT temp[9].xyz, temp[3].x___;
> 36: MUL temp[12].xy, temp[5].xy__, const[12].xy__;
> 37: TEX temp[3].xyz, temp[12].xy__, 2D[3];
> 38: MUL temp[3].xyz, temp[3].xyz_, const[7].xyz_;
> 39: TEX temp[2].xyz, temp[7].xyz_, CUBE[4];
> 40: ADD temp[5].xyz, temp[2].xyz_, -temp[3].xyz_;
> 41: MAD temp[5].xyz, temp[5].xyz_, temp[9].xyz_, temp[3].xyz_;
> 42: ADD output[0].xyz, temp[5].xyz_, temp[8].xyz_;
> 43: MUL output[0].w, input[1].___w, temp[4].___x;
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: RCP temp[10].w, input[0].w___;
>  1: MUL temp[10].xy, input[0].xy__, temp[10].ww__;
>  2: MAD temp[10].xy, temp[10].xy__, const[10].xy__, const[11].xy__;
>  3: MOV temp[0].x, temp[10].x___;
>  4: ADD temp[0].y, const[8]._y__, -temp[10]._y__;
>  5: TEX temp[1].yw, input[2].xy__, 2D[0];
>  6: MOV temp[1].x, temp[1].w___;
>  7: MAD temp[2].xy, temp[1].xy__, const[9].xx__, -none.11__;
>  8: MUL temp[2].xy, temp[2].xy__, const[1].zz__;
>  9: MOV temp[2].z, none.__0_;
> 10: DP3 temp[3].x, temp[2].xyz_, temp[2].xyz_;
> 11: ADD temp[3].x, none.1___, -temp[3].x___;
> 12: RSQ temp[3].x, |temp[3].x___|;
> 13: RCP temp[2].z, temp[3].x___;
> 14: TEX temp[4].x, input[2].xy__, 2D[1];
> 15: DP3 temp[3].x, temp[2].xyz_, input[4].xyz_;
> 16: DP3 temp[3].y, temp[2].xyz_, input[5].xyz_;
> 17: DP3 temp[3].z, temp[2].xyz_, input[6].xyz_;
> 18: MUL temp[5].xy, const[1].xy__, const[2].xy__;
> 19: MAD temp[5].xy, temp[5].xy__, temp[3].xy__, temp[0].xy__;
> 20: DP3 temp[6].w, input[3].xyz_, input[3].xyz_;
> 21: RSQ temp[6].w, |temp[6].w___|;
> 22: MUL temp[6].xyz, input[3].xyz_, temp[6].www_;
> 23: DP3 temp[2].x, temp[6].xyz_, temp[3].xyz_;
> 24: MUL temp[2].x, temp[2].x___, const[9].x___;
> 25: MAD temp[7].xyz, -temp[3].xyz_, temp[2].xxx_, temp[6].xyz_;
> 26: DP3_SAT temp[1].xy, temp[7].xyz_, const[3].xyz_;
> 27: TEX temp[8].xyz, temp[1].xy__, 2D[2];
> 28: MUL temp[8].xyz, temp[8].xyz_, const[4].xyz_;
> 29: MUL temp[8].xyz, temp[8].xyz_, const[5].xyz_;
> 30: DP3 temp[3].x, -temp[6].xyz_, temp[3].xyz_;
> 31: ADD temp[3].x, temp[3].x___, none.1___;
> 32: LG2 temp[11].w, temp[3].x___;
> 33: MUL temp[11].w, temp[11].___w, const[6].___x;
> 34: EX2 temp[3].x, temp[11].w___;
> 35: RCP_SAT temp[9].xyz, temp[3].x___;
> 36: MUL temp[12].xy, temp[5].xy__, const[12].xy__;
> 37: TEX temp[3].xyz, temp[12].xy__, 2D[3];
> 38: MUL temp[3].xyz, temp[3].xyz_, const[7].xyz_;
> 39: TEX temp[2].xyz, temp[7].xyz_, CUBE[4];
> 40: ADD temp[5].xyz, temp[2].xyz_, -temp[3].xyz_;
> 41: MAD temp[5].xyz, temp[5].xyz_, temp[9].xyz_, temp[3].xyz_;
> 42: ADD output[0].xyz, temp[5].xyz_, temp[8].xyz_;
> 43: MUL output[0].w, input[1].___w, temp[4].___x;
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: src0.w = input[0]
>     RCP temp[10].w, src0.w
>  1: src0.xyz = input[0], src0.w = temp[10]
>     MAD temp[10].xy, src0.xy_, src0.ww_, src0.000
>  2: src0.xyz = temp[10], src1.xyz = const[10], src2.xyz = const[11]
>     MAD temp[10].xy, src0.xy_, src1.xy_, src2.xy_
>  3: src0.xyz = temp[10]
>     MAD temp[0].x, src0.x__, src0.111, src0.000
>  4: src0.xyz = const[8], src1.xyz = temp[10]
>     MAD temp[0].y, src0._y_, src0.111, -src1._y_
>  5: TEX temp[1].yw, input[2].xy__, 2D[0];
>  6: src0.w = temp[1]
>     MAD temp[1].x, src0.w__, src0.111, src0.000
>  7: src0.xyz = temp[1], src1.xyz = const[9]
>     MAD temp[2].xy, src0.xy_, src1.xx_, -src0.11_
>  8: src0.xyz = temp[2], src1.xyz = const[1]
>     MAD temp[2].xy, src0.xy_, src1.zz_, src0.000
>  9: 
>     MAD temp[2].z, src0.__0, src0.111, src0.000
> 10: src0.xyz = temp[2]
>     DP3 temp[3].x, src0.xyz, src0.xyz
> 11: src0.xyz = temp[3]
>     MAD temp[3].x, src0.1__, src0.111, -src0.x__
> 12: src0.xyz = temp[3]
>     REPL_ALPHA temp[3].x
>     RSQ, |src0.x|
> 13: src0.xyz = temp[3]
>     REPL_ALPHA temp[2].z
>     RCP, src0.x
> 14: TEX temp[4].x, input[2].xy__, 2D[1];
> 15: src0.xyz = temp[2], src1.xyz = input[4]
>     DP3 temp[3].x, src0.xyz, src1.xyz
> 16: src0.xyz = temp[2], src1.xyz = input[5]
>     DP3 temp[3].y, src0.xyz, src1.xyz
> 17: src0.xyz = temp[2], src1.xyz = input[6]
>     DP3 temp[3].z, src0.xyz, src1.xyz
> 18: src0.xyz = const[1], src1.xyz = const[2]
>     MAD temp[5].xy, src0.xy_, src1.xy_, src0.000
> 19: src0.xyz = temp[5], src1.xyz = temp[3], src2.xyz = temp[0]
>     MAD temp[5].xy, src0.xy_, src1.xy_, src2.xy_
> 20: src0.xyz = input[3]
>     DP3, src0.xyz, src0.xyz
>     DP3 temp[6].w, src0._, src0._
> 21: src0.w = temp[6]
>     RSQ temp[6].w, |src0.w|
> 22: src0.xyz = input[3], src0.w = temp[6]
>     MAD temp[6].xyz, src0.xyz, src0.www, src0.000
> 23: src0.xyz = temp[6], src1.xyz = temp[3]
>     DP3 temp[2].x, src0.xyz, src1.xyz
> 24: src0.xyz = temp[2], src1.xyz = const[9]
>     MAD temp[2].x, src0.x__, src1.x__, src0.000
> 25: src0.xyz = temp[3], src1.xyz = temp[2], src2.xyz = temp[6]
>     MAD temp[7].xyz, -src0.xyz, src1.xxx, src2.xyz
> 26: src0.xyz = temp[7], src1.xyz = const[3]
>     DP3_SAT temp[1].xy, src0.xyz, src1.xyz
> 27: TEX temp[8].xyz, temp[1].xy__, 2D[2];
> 28: src0.xyz = temp[8], src1.xyz = const[4]
>     MAD temp[8].xyz, src0.xyz, src1.xyz, src0.000
> 29: src0.xyz = temp[8], src1.xyz = const[5]
>     MAD temp[8].xyz, src0.xyz, src1.xyz, src0.000
> 30: src0.xyz = temp[6], src1.xyz = temp[3]
>     DP3 temp[3].x, -src0.xyz, src1.xyz
> 31: src0.xyz = temp[3]
>     MAD temp[3].x, src0.x__, src0.111, src0.1__
> 32: src0.xyz = temp[3]
>     LG2 temp[11].w, src0.x
> 33: src0.xyz = const[6], src0.w = temp[11]
>     MAD temp[11].w, src0.w, src0.x, src0.0
> 34: src0.w = temp[11]
>     REPL_ALPHA temp[3].x
>     EX2, src0.w
> 35: src0.xyz = temp[3]
>     REPL_ALPHA_SAT temp[9].xyz
>     RCP_SAT, src0.x
> 36: src0.xyz = temp[5], src1.xyz = const[12]
>     MAD temp[12].xy, src0.xy_, src1.xy_, src0.000
> 37: TEX temp[3].xyz, temp[12].xy__, 2D[3];
> 38: src0.xyz = temp[3], src1.xyz = const[7]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src0.000
> 39: TEX temp[2].xyz, temp[7].xyz_, CUBE[4];
> 40: src0.xyz = temp[2], src1.xyz = temp[3]
>     MAD temp[5].xyz, src0.xyz, src0.111, -src1.xyz
> 41: src0.xyz = temp[5], src1.xyz = temp[9], src2.xyz = temp[3]
>     MAD temp[5].xyz, src0.xyz, src1.xyz, src2.xyz
> 42: src0.xyz = temp[5], src1.xyz = temp[8]
>     MAD color[0].xyz, src0.xyz, src0.111, src1.xyz
> 43: src0.xyz = temp[4], src0.w = input[1]
>     MAD color[0].w, src0.w, src0.x, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[4].x, input[2].xy__, 2D[1];
>  2: TEX temp[1].yw, input[2].xy__, 2D[0];
>  3: src0.xyz = input[3]
>     DP3, src0.xyz, src0.xyz
>     DP3 temp[6].w, src0._, src0._
>  4: src0.w = temp[1], src1.w = temp[6]
>     MAD temp[1].x, src0.w__, src0.111, src0.000
>     RSQ temp[6].w, |src1.w|
>  5: src0.xyz = input[3], src0.w = temp[6], src1.xyz = temp[4], src1.w = input[1]
>     MAD temp[6].xyz, src0.xyz, src0.www, src0.000
>     MAD color[0].w, src1.w, src1.x, src0.0
>  6: src0.xyz = temp[1], src0.w = input[0], src1.xyz = const[9]
>     MAD temp[2].xy, src0.xy_, src1.xx_, -src0.11_
>     RCP temp[10].w, src0.w
>  7: src0.xyz = input[0], src0.w = temp[10]
>     MAD temp[10].xy, src0.xy_, src0.ww_, src0.000
>  8: src0.xyz = temp[10], src1.xyz = const[10], src2.xyz = const[11]
>     MAD temp[10].xy, src0.xy_, src1.xy_, src2.xy_
>  9: src0.xyz = const[8], src1.xyz = temp[10]
>     MAD temp[0].y, src0._y_, src0.111, -src1._y_
> 10: src0.xyz = temp[10]
>     MAD temp[0].x, src0.x__, src0.111, src0.000
> 11: src0.xyz = temp[2], src1.xyz = const[1]
>     MAD temp[2].xy, src0.xy_, src1.zz_, src0.000
> 12: src0.xyz = const[1], src1.xyz = const[2]
>     MAD temp[5].xy, src0.xy_, src1.xy_, src0.000
> 13: 
>     MAD temp[2].z, src0.__0, src0.111, src0.000
> 14: src0.xyz = temp[2]
>     DP3 temp[3].x, src0.xyz, src0.xyz
> 15: src0.xyz = temp[3]
>     MAD temp[3].x, src0.1__, src0.111, -src0.x__
> 16: src0.xyz = temp[3]
>     REPL_ALPHA temp[3].x
>     RSQ, |src0.x|
> 17: src0.xyz = temp[3]
>     REPL_ALPHA temp[2].z
>     RCP, src0.x
> 18: src0.xyz = temp[2], src1.xyz = input[4]
>     DP3 temp[3].x, src0.xyz, src1.xyz
> 19: src0.xyz = temp[2], src1.xyz = input[5]
>     DP3 temp[3].y, src0.xyz, src1.xyz
> 20: src0.xyz = temp[5], src1.xyz = temp[3], src2.xyz = temp[0]
>     MAD temp[5].xy, src0.xy_, src1.xy_, src2.xy_
> 21: src0.xyz = temp[5], src1.xyz = const[12]
>     MAD temp[12].xy, src0.xy_, src1.xy_, src0.000
> 22: src0.xyz = temp[2], src1.xyz = input[6]
>     DP3 temp[3].z, src0.xyz, src1.xyz
> 23: src0.xyz = temp[6], src1.xyz = temp[3]
>     DP3 temp[2].x, src0.xyz, src1.xyz
> 24: src0.xyz = temp[2], src1.xyz = const[9]
>     MAD temp[2].x, src0.x__, src1.x__, src0.000
> 25: src0.xyz = temp[3], src1.xyz = temp[2], src2.xyz = temp[6]
>     MAD temp[7].xyz, -src0.xyz, src1.xxx, src2.xyz
> 26: src0.xyz = temp[7], src1.xyz = const[3]
>     DP3_SAT temp[1].xy, src0.xyz, src1.xyz
> 27: src0.xyz = temp[6], src1.xyz = temp[3]
>     DP3 temp[3].x, -src0.xyz, src1.xyz
> 28: src0.xyz = temp[3]
>     MAD temp[3].x, src0.x__, src0.111, src0.1__
> 29: src0.xyz = temp[3]
>     LG2 temp[11].w, src0.x
> 30: src0.xyz = const[6], src0.w = temp[11]
>     MAD temp[11].w, src0.w, src0.x, src0.0
> 31: src0.w = temp[11]
>     REPL_ALPHA temp[3].x
>     EX2, src0.w
> 32: src0.xyz = temp[3]
>     REPL_ALPHA_SAT temp[9].xyz
>     RCP_SAT, src0.x
> 33: BEGIN_TEX;
> 34: TEX temp[3].xyz, temp[12].xy__, 2D[3];
> 35: TEX temp[8].xyz, temp[1].xy__, 2D[2];
> 36: TEX temp[2].xyz, temp[7].xyz_, CUBE[4];
> 37: src0.xyz = temp[8], src1.xyz = const[4]
>     MAD temp[8].xyz, src0.xyz, src1.xyz, src0.000
> 38: src0.xyz = temp[8], src1.xyz = const[5]
>     MAD temp[8].xyz, src0.xyz, src1.xyz, src0.000
> 39: src0.xyz = temp[3], src1.xyz = const[7]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src0.000
> 40: src0.xyz = temp[2], src1.xyz = temp[3]
>     MAD temp[5].xyz, src0.xyz, src0.111, -src1.xyz
> 41: src0.xyz = temp[5], src1.xyz = temp[9], src2.xyz = temp[3]
>     MAD temp[5].xyz, src0.xyz, src1.xyz, src2.xyz
> 42: src0.xyz = temp[5], src1.xyz = temp[8]
>     MAD color[0].xyz, src0.xyz, src0.111, src1.xyz
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[7].x, temp[1].xy__, 2D[1];
>  2: TEX temp[1].yw, temp[1].xy__, 2D[0];
>  3: src0.xyz = temp[2]
>     DP3, src0.xyz, src0.xyz
>     DP3 temp[8].w, src0._, src0._
>  4: src0.w = temp[1], src1.w = temp[8]
>     MAD temp[1].x, src0.w__, src0.111, src0.000
>     RSQ temp[8].w, |src1.w|
>  5: src0.xyz = temp[2], src0.w = temp[8], src1.xyz = temp[7], src1.w = temp[0]
>     MAD temp[8].xyz, src0.xyz, src0.www, src0.000
>     MAD color[0].w, src1.w, src1.x, src0.0
>  6: src0.xyz = temp[1], src0.w = temp[6], src1.xyz = const[9]
>     MAD temp[2].xy, src0.xy_, src1.xx_, -src0.11_
>     RCP temp[7].w, src0.w
>  7: src0.xyz = temp[6], src0.w = temp[7]
>     MAD temp[7].xy, src0.xy_, src0.ww_, src0.000
>  8: src0.xyz = temp[7], src1.xyz = const[10], src2.xyz = const[11]
>     MAD temp[7].xy, src0.xy_, src1.xy_, src2.xy_
>  9: src0.xyz = const[8], src1.xyz = temp[7]
>     MAD temp[0].y, src0._y_, src0.111, -src1._y_
> 10: src0.xyz = temp[7]
>     MAD temp[0].x, src0.x__, src0.111, src0.000
> 11: src0.xyz = temp[2], src1.xyz = const[1]
>     MAD temp[2].xy, src0.xy_, src1.zz_, src0.000
> 12: src0.xyz = const[1], src1.xyz = const[2]
>     MAD temp[7].xy, src0.xy_, src1.xy_, src0.000
> 13: 
>     MAD temp[2].z, src0.__0, src0.111, src0.000
> 14: src0.xyz = temp[2]
>     DP3 temp[6].x, src0.xyz, src0.xyz
> 15: src0.xyz = temp[6]
>     MAD temp[6].x, src0.1__, src0.111, -src0.x__
> 16: src0.xyz = temp[6]
>     REPL_ALPHA temp[6].x
>     RSQ, |src0.x|
> 17: src0.xyz = temp[6]
>     REPL_ALPHA temp[2].z
>     RCP, src0.x
> 18: src0.xyz = temp[2], src1.xyz = temp[3]
>     DP3 temp[6].x, src0.xyz, src1.xyz
> 19: src0.xyz = temp[2], src1.xyz = temp[4]
>     DP3 temp[6].y, src0.xyz, src1.xyz
> 20: src0.xyz = temp[7], src1.xyz = temp[6], src2.xyz = temp[0]
>     MAD temp[7].xy, src0.xy_, src1.xy_, src2.xy_
> 21: src0.xyz = temp[7], src1.xyz = const[12]
>     MAD temp[4].xy, src0.xy_, src1.xy_, src0.000
> 22: src0.xyz = temp[2], src1.xyz = temp[5]
>     DP3 temp[6].z, src0.xyz, src1.xyz
> 23: src0.xyz = temp[8], src1.xyz = temp[6]
>     DP3 temp[2].x, src0.xyz, src1.xyz
> 24: src0.xyz = temp[2], src1.xyz = const[9]
>     MAD temp[2].x, src0.x__, src1.x__, src0.000
> 25: src0.xyz = temp[6], src1.xyz = temp[2], src2.xyz = temp[8]
>     MAD temp[0].xyz, -src0.xyz, src1.xxx, src2.xyz
> 26: src0.xyz = temp[0], src1.xyz = const[3]
>     DP3_SAT temp[1].xy, src0.xyz, src1.xyz
> 27: src0.xyz = temp[8], src1.xyz = temp[6]
>     DP3 temp[6].x, -src0.xyz, src1.xyz
> 28: src0.xyz = temp[6]
>     MAD temp[6].x, src0.x__, src0.111, src0.1__
> 29: src0.xyz = temp[6]
>     LG2 temp[3].w, src0.x
> 30: src0.xyz = const[6], src0.w = temp[3]
>     MAD temp[3].w, src0.w, src0.x, src0.0
> 31: src0.w = temp[3]
>     REPL_ALPHA temp[6].x
>     EX2, src0.w
> 32: src0.xyz = temp[6]
>     REPL_ALPHA_SAT temp[3].xyz
>     RCP_SAT, src0.x
> 33: BEGIN_TEX;
> 34: TEX temp[6].xyz, temp[4].xy__, 2D[3];
> 35: TEX temp[1].xyz, temp[1].xy__, 2D[2];
> 36: TEX temp[2].xyz, temp[0].xyz_, CUBE[4];
> 37: src0.xyz = temp[1], src1.xyz = const[4]
>     MAD temp[1].xyz, src0.xyz, src1.xyz, src0.000
> 38: src0.xyz = temp[1], src1.xyz = const[5]
>     MAD temp[1].xyz, src0.xyz, src1.xyz, src0.000
> 39: src0.xyz = temp[6], src1.xyz = const[7]
>     MAD temp[6].xyz, src0.xyz, src1.xyz, src0.000
> 40: src0.xyz = temp[2], src1.xyz = temp[6]
>     MAD temp[7].xyz, src0.xyz, src0.111, -src1.xyz
> 41: src0.xyz = temp[7], src1.xyz = temp[3], src2.xyz = temp[6]
>     MAD temp[7].xyz, src0.xyz, src1.xyz, src2.xyz
> 42: src0.xyz = temp[7], src1.xyz = temp[1]
>     MAD color[0].xyz, src0.xyz, src0.111, src1.xyz
>pc=52*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 29, tex_end: 1  (code_addr: 00420740)
>  TEX:
>    TEX t7, t1, texture[1]   (000089c1)
>    TEX t1, t1, texture[0]   (00008041)
>  0: xyz:  t2  t0  t0 ->                      (00000002)
>       w:  t0  t0  t0 -> t8.w                 (00a00000)
>     xyz:   t2.xyz   t2.xyz   t2.xxx    op: 00804000
>       w:      1.0      1.0     t2.x    op: 00800891
>  1: xyz:  t0  t0  t0 -> t1.x                 (00840000)
>       w:  t1  t8  t0 -> t8.w                 (00a00201)
>     xyz:   t1.www      1.0      0.0    op: 00050a8c
>       w:   |t8.w|     t0.x     t0.x    op: 0580004a
>  2: xyz:  t2  t7  t0 -> t8.xyz               (03a001c2)
>       w:  t8  t0  t0 -> o0.w                 (01000008)
>     xyz:   t2.xyz   t8.www      0.0    op: 00050600
>       w:     t0.w     t7.x      0.0    op: 0004018a
>  3: xyz:  t1  c9  t0 -> t2.xy                (01880a41)
>       w:  t6  t0  t0 -> t7.w                 (009c0006)
>     xyz:   t1.xyz   c9.xxx     -1.0    op: 000d4280
>       w:     t6.w     t1.x     t1.x    op: 05000009
>  4: xyz:  t6  t0  t0 -> t7.xy                (019c0006)
>       w:  t7  t0  t0 ->                      (00000007)
>     xyz:   t6.xyz   t7.www      0.0    op: 00050600
>       w:     t6.x     t6.x     t6.x    op: 00000000
>  5: xyz:  t7 c10 c11 -> t7.xy                (019eba87)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t7.xyz  c10.xyz  c11.xyz    op: 00020200
>       w:     t7.x     t7.x     t7.x    op: 00000000
>  6: xyz:  c8  t7  t0 -> t0.y                 (010001e8)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   c8.xyz      1.0  -t7.xyz    op: 00090a80
>       w:     c8.x     c8.x     c8.x    op: 00000000
>  7: xyz:  t7  t0  t0 -> t0.x                 (00800007)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t7.xyz      1.0      0.0    op: 00050a80
>       w:     t7.x     t7.x     t7.x    op: 00000000
>  8: xyz:  t2  c1  t0 -> t2.xy                (01880842)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz   c1.zzz      0.0    op: 00050380
>       w:     t2.x     t2.x     t2.x    op: 00000000
>  9: xyz:  c1  c2  t0 -> t7.xy                (019c08a1)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   c1.xyz   c2.xyz      0.0    op: 00050200
>       w:     c1.x     c1.x     c1.x    op: 00000000
> 10: xyz:  t0  t0  t0 -> t2.z                 (02080000)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:      0.0      1.0      0.0    op: 00050a94
>       w:     t0.x     t0.x     t0.x    op: 00000000
> 11: xyz:  t2  t0  t0 -> t6.x                 (00980002)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz   t2.xyz   t2.xxx    op: 00804000
>       w:     t2.x     t2.x     t2.x    op: 00000000
> 12: xyz:  t6  t0  t0 -> t6.x                 (00980006)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:      1.0      1.0  -t6.xyz    op: 00080a95
>       w:     t6.x     t6.x     t6.x    op: 00000000
> 13: xyz:  t6  t0  t0 -> t6.x                 (00980006)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t6.xxx   t6.xxx   t6.xxx    op: 05004081
>       w:   |t6.x|     t6.x     t6.x    op: 05800040
> 14: xyz:  t6  t0  t0 -> t2.z                 (02080006)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t6.xxx   t6.xxx   t6.xxx    op: 05004081
>       w:     t6.x     t6.x     t6.x    op: 05000000
> 15: xyz:  t2  t3  t0 -> t6.x                 (009800c2)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz   t3.xyz   t2.xxx    op: 00804200
>       w:     t2.x     t2.x     t2.x    op: 00000000
> 16: xyz:  t2  t4  t0 -> t6.y                 (01180102)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz   t4.xyz   t2.xxx    op: 00804200
>       w:     t2.x     t2.x     t2.x    op: 00000000
> 17: xyz:  t7  t6  t0 -> t7.xy                (019c0187)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t7.xyz   t6.xyz   t0.xyz    op: 00020200
>       w:     t7.x     t7.x     t7.x    op: 00000000
> 18: xyz:  t7 c12  t0 -> t4.xy                (01900b07)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t7.xyz  c12.xyz      0.0    op: 00050200
>       w:     t7.x     t7.x     t7.x    op: 00000000
> 19: xyz:  t2  t5  t0 -> t6.z                 (02180142)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz   t5.xyz   t2.xxx    op: 00804200
>       w:     t2.x     t2.x     t2.x    op: 00000000
> 20: xyz:  t8  t6  t0 -> t2.x                 (00880188)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t8.xyz   t6.xyz   t8.xxx    op: 00804200
>       w:     t8.x     t8.x     t8.x    op: 00000000
> 21: xyz:  t2  c9  t0 -> t2.x                 (00880a42)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz   c9.xyz      0.0    op: 00050200
>       w:     t2.x     t2.x     t2.x    op: 00000000
> 22: xyz:  t6  t2  t8 -> t0.xyz               (03808086)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:  -t6.xyz   t2.xxx   t8.xyz    op: 000202a0
>       w:     t6.x     t6.x     t6.x    op: 00000000
> 23: xyz:  t0  c3  t0 -> t1.xy                (018408c0)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xyz   c3.xyz   t0.xxx    op: 40804200
>       w:     t0.x     t0.x     t0.x    op: 00000000
> 24: xyz:  t8  t6  t0 -> t6.x                 (00980188)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:  -t8.xyz   t6.xyz   t8.xxx    op: 00804220
>       w:     t8.x     t8.x     t8.x    op: 00000000
> 25: xyz:  t6  t0  t0 -> t6.x                 (00980006)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t6.xyz      1.0      1.0    op: 00054a80
>       w:     t6.x     t6.x     t6.x    op: 00000000
> 26: xyz:  t6  t0  t0 ->                      (00000006)
>       w:  t0  t0  t0 -> t3.w                 (008c0000)
>     xyz:   t6.xxx   t6.xxx   t6.xxx    op: 00004081
>       w:     t6.x     t6.x     t6.x    op: 04800000
> 27: xyz:  c6  t0  t0 ->                      (00000026)
>       w:  t3  t0  t0 -> t3.w                 (008c0003)
>     xyz:   c6.xxx   c6.xxx   c6.xxx    op: 00004081
>       w:     t3.w     c6.x      0.0    op: 00040009
> 28: xyz:  t0  t0  t0 -> t6.x                 (00980000)
>       w:  t3  t0  t0 ->                      (00000003)
>     xyz:   t0.xxx   t0.xxx   t0.xxx    op: 05004081
>       w:     t3.w     t0.x     t0.x    op: 04000009
> 29: xyz:  t6  t0  t0 -> t3.xyz               (038c0006)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t6.xxx   t6.xxx   t6.xxx    op: 45004081
>       w:     t6.x     t6.x     t6.x    op: 45000000
>NODE 1: alu_offset: 30, tex_offset: 2, alu_end: 5, tex_end: 2  (code_addr: 0044215e)
>  TEX:
>    TEX t6, t4, texture[3]   (00009984)
>    TEX t1, t1, texture[2]   (00009041)
>    TEX t2, t0, texture[4]   (0000a080)
> 30: xyz:  t1  c4  t0 -> t1.xyz               (03840901)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   c4.xyz      0.0    op: 00050200
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 31: xyz:  t1  c5  t0 -> t1.xyz               (03840941)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   c5.xyz      0.0    op: 00050200
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 32: xyz:  t6  c7  t0 -> t6.xyz               (039809c6)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t6.xyz   c7.xyz      0.0    op: 00050200
>       w:     t6.x     t6.x     t6.x    op: 00000000
> 33: xyz:  t2  t6  t0 -> t7.xyz               (039c0182)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz      1.0  -t6.xyz    op: 00090a80
>       w:     t2.x     t2.x     t2.x    op: 00000000
> 34: xyz:  t7  t3  t6 -> t7.xyz               (039c60c7)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t7.xyz   t3.xyz   t6.xyz    op: 00020200
>       w:     t7.x     t7.x     t7.x    op: 00000000
> 35: xyz:  t7  t1  t0 -> o0.xyz               (1c000047)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t7.xyz      1.0   t1.xyz    op: 00010a80
>       w:     t7.x     t7.x     t7.x    op: 00000000
>r300: Initial fragment program
>FRAG
>DCL IN[0], COLOR, LINEAR
>DCL IN[1], GENERIC[0], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL TEMP[0]
>  0: TEX TEMP[0], IN[1], SAMP[0], 2D
>  1: MUL OUT[0], TEMP[0], IN[0]
>  2: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: MUL output[0], temp[0], input[0];
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: MUL output[0], temp[0], input[0];
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: MUL output[0], temp[0], input[0];
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: MUL output[0], temp[0], input[0];
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: TEX temp[0], input[1], 2D[0];
>  1: MUL output[0], temp[0], input[0];
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: TEX temp[0], input[1].xy__, 2D[0];
>  1: MUL output[0], temp[0], input[0];
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: TEX temp[0], input[1].xy__, 2D[0];
>  1: MUL output[0], temp[0], input[0];
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: TEX temp[0], input[1].xy__, 2D[0];
>  1: MUL output[0], temp[0], input[0];
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: TEX temp[0], input[1].xy__, 2D[0];
>  1: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = input[0], src1.w = input[0]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[0], input[1].xy__, 2D[0];
>  2: src0.xyz = temp[0], src0.w = temp[0], src1.xyz = input[0], src1.w = input[0]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[1], temp[1].xy__, 2D[0];
>  2: src0.xyz = temp[1], src0.w = temp[1], src1.xyz = temp[0], src1.w = temp[0]
>     MAD color[0].xyz, src0.xyz, src1.xyz, src0.000
>     MAD color[0].w, src0.w, src1.w, src0.0
>pc=53*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 0, tex_end: 0  (code_addr: 00400000)
>  TEX:
>    TEX t1, t1, texture[0]   (00008041)
>  0: xyz:  t1  t0  t0 -> o0.xyz               (1c000001)
>       w:  t1  t0  t0 -> o0.w                 (01000001)
>     xyz:   t1.xyz   t0.xyz      0.0    op: 00050200
>       w:     t1.w     t0.w      0.0    op: 00040509
>r300: Initial fragment program
>FRAG
>DCL IN[0], POSITION, LINEAR
>DCL IN[1], COLOR, LINEAR
>DCL IN[2], GENERIC[0], PERSPECTIVE
>DCL IN[3], GENERIC[2], PERSPECTIVE
>DCL IN[4], GENERIC[3], PERSPECTIVE
>DCL IN[5], GENERIC[4], PERSPECTIVE
>DCL IN[6], GENERIC[5], PERSPECTIVE
>DCL OUT[0], COLOR
>DCL SAMP[0]
>DCL SAMP[1]
>DCL SAMP[2]
>DCL SAMP[3]
>DCL SAMP[4]
>DCL CONST[7..8]
>DCL CONST[1..6]
>DCL TEMP[0..9]
>IMM FLT32 {    2.0000,     1.0000,     0.0000,     0.0000}
>  0: MOV TEMP[0], IN[0]
>  1: SUB TEMP[0].y, CONST[8], IN[0]
>  2: TEX TEMP[1], IN[2], SAMP[0], 2D
>  3: MOV TEMP[1].x, TEMP[1].wwww
>  4: MAD TEMP[2], TEMP[1], IMM[0].xxxx, -IMM[0].yyyy
>  5: MUL TEMP[2].xy, TEMP[2], CONST[1].zzzz
>  6: MOV TEMP[2].z, IMM[0].zzzz
>  7: DP3 TEMP[3].x, TEMP[2], TEMP[2]
>  8: ADD TEMP[3].x, IMM[0].yyyy, -TEMP[3].xxxx
>  9: RSQ TEMP[3].x, TEMP[3].xxxx
> 10: RCP TEMP[2].z, TEMP[3].xxxx
> 11: TEX TEMP[4], IN[2], SAMP[1], 2D
> 12: DP3 TEMP[3].x, TEMP[2], IN[4]
> 13: DP3 TEMP[3].y, TEMP[2], IN[5]
> 14: DP3 TEMP[3].z, TEMP[2], IN[6]
> 15: MUL TEMP[5], CONST[1], CONST[2]
> 16: MAD TEMP[5], TEMP[5], TEMP[3], TEMP[0]
> 17: DP3 TEMP[6].w, IN[3], IN[3]
> 18: RSQ TEMP[6].w, TEMP[6].wwww
> 19: MUL TEMP[6], IN[3], TEMP[6].wwww
> 20: DP3 TEMP[2], TEMP[6], TEMP[3]
> 21: MUL TEMP[2], TEMP[2], IMM[0].xxxx
> 22: MAD TEMP[7], -TEMP[3], TEMP[2].xxxx, TEMP[6]
> 23: DP3_SAT TEMP[1], TEMP[7], CONST[3]
> 24: TEX TEMP[8], TEMP[1], SAMP[2], 2D
> 25: MUL TEMP[8], TEMP[8], CONST[4]
> 26: MUL TEMP[8], TEMP[8], CONST[5]
> 27: DP3 TEMP[3], -TEMP[6], TEMP[3]
> 28: ADD TEMP[3], TEMP[3], IMM[0].yyyy
> 29: POW TEMP[3], TEMP[3].xxxx, CONST[6].xxxx
> 30: RCP_SAT TEMP[9], TEMP[3].xxxx
> 31: TEX TEMP[3], TEMP[5], SAMP[3], RECT
> 32: MUL TEMP[3], TEMP[3], CONST[7]
> 33: TEX TEMP[2], TEMP[7], SAMP[4], CUBE
> 34: SUB TEMP[5], TEMP[2], TEMP[3]
> 35: MAD TEMP[5], TEMP[5], TEMP[9], TEMP[3]
> 36: ADD OUT[0], TEMP[5], TEMP[8]
> 37: MUL OUT[0].w, IN[1].wwww, TEMP[4].xxxx
> 38: END
>Fragment Program: before compilation
># Radeon Compiler Program
>  0: RCP temp[10].w, input[0].wwww;
>  1: MUL temp[10].xyz, input[0], temp[10].wwww;
>  2: MAD temp[10].xyz, temp[10].xyz0, const[10].xyz0, const[11].xyz0;
>  3: MOV temp[0], temp[10];
>  4: SUB temp[0].y, const[8], temp[10];
>  5: TEX temp[1], input[2], 2D[0];
>  6: MOV temp[1].x, temp[1].wwww;
>  7: MAD temp[2], temp[1], const[9].xxxx, -const[9].yyyy;
>  8: MUL temp[2].xy, temp[2], const[1].zzzz;
>  9: MOV temp[2].z, const[9].zzzz;
> 10: DP3 temp[3].x, temp[2], temp[2];
> 11: ADD temp[3].x, const[9].yyyy, -temp[3].xxxx;
> 12: RSQ temp[3].x, temp[3].xxxx;
> 13: RCP temp[2].z, temp[3].xxxx;
> 14: TEX temp[4], input[2], 2D[1];
> 15: DP3 temp[3].x, temp[2], input[4];
> 16: DP3 temp[3].y, temp[2], input[5];
> 17: DP3 temp[3].z, temp[2], input[6];
> 18: MUL temp[5], const[1], const[2];
> 19: MAD temp[5], temp[5], temp[3], temp[0];
> 20: DP3 temp[6].w, input[3], input[3];
> 21: RSQ temp[6].w, temp[6].wwww;
> 22: MUL temp[6], input[3], temp[6].wwww;
> 23: DP3 temp[2], temp[6], temp[3];
> 24: MUL temp[2], temp[2], const[9].xxxx;
> 25: MAD temp[7], -temp[3], temp[2].xxxx, temp[6];
> 26: DP3_SAT temp[1], temp[7], const[3];
> 27: TEX temp[8], temp[1], 2D[2];
> 28: MUL temp[8], temp[8], const[4];
> 29: MUL temp[8], temp[8], const[5];
> 30: DP3 temp[3], -temp[6], temp[3];
> 31: ADD temp[3], temp[3], const[9].yyyy;
> 32: POW temp[3], temp[3].xxxx, const[6].xxxx;
> 33: RCP_SAT temp[9], temp[3].xxxx;
> 34: TEX temp[3], temp[5], RECT[3];
> 35: MUL temp[3], temp[3], const[7];
> 36: TEX temp[2], temp[7], CUBE[4];
> 37: SUB temp[5], temp[2], temp[3];
> 38: MAD temp[5], temp[5], temp[9], temp[3];
> 39: ADD output[0], temp[5], temp[8];
> 40: MUL output[0].w, input[1].wwww, temp[4].xxxx;
>Fragment Program: after emulate loops
># Radeon Compiler Program
>  0: RCP temp[10].w, input[0].wwww;
>  1: MUL temp[10].xyz, input[0], temp[10].wwww;
>  2: MAD temp[10].xyz, temp[10].xyz0, const[10].xyz0, const[11].xyz0;
>  3: MOV temp[0], temp[10];
>  4: SUB temp[0].y, const[8], temp[10];
>  5: TEX temp[1], input[2], 2D[0];
>  6: MOV temp[1].x, temp[1].wwww;
>  7: MAD temp[2], temp[1], const[9].xxxx, -const[9].yyyy;
>  8: MUL temp[2].xy, temp[2], const[1].zzzz;
>  9: MOV temp[2].z, const[9].zzzz;
> 10: DP3 temp[3].x, temp[2], temp[2];
> 11: ADD temp[3].x, const[9].yyyy, -temp[3].xxxx;
> 12: RSQ temp[3].x, temp[3].xxxx;
> 13: RCP temp[2].z, temp[3].xxxx;
> 14: TEX temp[4], input[2], 2D[1];
> 15: DP3 temp[3].x, temp[2], input[4];
> 16: DP3 temp[3].y, temp[2], input[5];
> 17: DP3 temp[3].z, temp[2], input[6];
> 18: MUL temp[5], const[1], const[2];
> 19: MAD temp[5], temp[5], temp[3], temp[0];
> 20: DP3 temp[6].w, input[3], input[3];
> 21: RSQ temp[6].w, temp[6].wwww;
> 22: MUL temp[6], input[3], temp[6].wwww;
> 23: DP3 temp[2], temp[6], temp[3];
> 24: MUL temp[2], temp[2], const[9].xxxx;
> 25: MAD temp[7], -temp[3], temp[2].xxxx, temp[6];
> 26: DP3_SAT temp[1], temp[7], const[3];
> 27: TEX temp[8], temp[1], 2D[2];
> 28: MUL temp[8], temp[8], const[4];
> 29: MUL temp[8], temp[8], const[5];
> 30: DP3 temp[3], -temp[6], temp[3];
> 31: ADD temp[3], temp[3], const[9].yyyy;
> 32: POW temp[3], temp[3].xxxx, const[6].xxxx;
> 33: RCP_SAT temp[9], temp[3].xxxx;
> 34: TEX temp[3], temp[5], RECT[3];
> 35: MUL temp[3], temp[3], const[7];
> 36: TEX temp[2], temp[7], CUBE[4];
> 37: SUB temp[5], temp[2], temp[3];
> 38: MAD temp[5], temp[5], temp[9], temp[3];
> 39: ADD output[0], temp[5], temp[8];
> 40: MUL output[0].w, input[1].wwww, temp[4].xxxx;
>Fragment Program: after emulate branches
># Radeon Compiler Program
>  0: RCP temp[10].w, input[0].wwww;
>  1: MUL temp[10].xyz, input[0], temp[10].wwww;
>  2: MAD temp[10].xyz, temp[10].xyz0, const[10].xyz0, const[11].xyz0;
>  3: MOV temp[0], temp[10];
>  4: SUB temp[0].y, const[8], temp[10];
>  5: TEX temp[1], input[2], 2D[0];
>  6: MOV temp[1].x, temp[1].wwww;
>  7: MAD temp[2], temp[1], const[9].xxxx, -const[9].yyyy;
>  8: MUL temp[2].xy, temp[2], const[1].zzzz;
>  9: MOV temp[2].z, const[9].zzzz;
> 10: DP3 temp[3].x, temp[2], temp[2];
> 11: ADD temp[3].x, const[9].yyyy, -temp[3].xxxx;
> 12: RSQ temp[3].x, temp[3].xxxx;
> 13: RCP temp[2].z, temp[3].xxxx;
> 14: TEX temp[4], input[2], 2D[1];
> 15: DP3 temp[3].x, temp[2], input[4];
> 16: DP3 temp[3].y, temp[2], input[5];
> 17: DP3 temp[3].z, temp[2], input[6];
> 18: MUL temp[5], const[1], const[2];
> 19: MAD temp[5], temp[5], temp[3], temp[0];
> 20: DP3 temp[6].w, input[3], input[3];
> 21: RSQ temp[6].w, temp[6].wwww;
> 22: MUL temp[6], input[3], temp[6].wwww;
> 23: DP3 temp[2], temp[6], temp[3];
> 24: MUL temp[2], temp[2], const[9].xxxx;
> 25: MAD temp[7], -temp[3], temp[2].xxxx, temp[6];
> 26: DP3_SAT temp[1], temp[7], const[3];
> 27: TEX temp[8], temp[1], 2D[2];
> 28: MUL temp[8], temp[8], const[4];
> 29: MUL temp[8], temp[8], const[5];
> 30: DP3 temp[3], -temp[6], temp[3];
> 31: ADD temp[3], temp[3], const[9].yyyy;
> 32: POW temp[3], temp[3].xxxx, const[6].xxxx;
> 33: RCP_SAT temp[9], temp[3].xxxx;
> 34: TEX temp[3], temp[5], RECT[3];
> 35: MUL temp[3], temp[3], const[7];
> 36: TEX temp[2], temp[7], CUBE[4];
> 37: SUB temp[5], temp[2], temp[3];
> 38: MAD temp[5], temp[5], temp[9], temp[3];
> 39: ADD output[0], temp[5], temp[8];
> 40: MUL output[0].w, input[1].wwww, temp[4].xxxx;
>Fragment Program: after native rewrite part 1
># Radeon Compiler Program
>  0: RCP temp[10].w, input[0].wwww;
>  1: MUL temp[10].xyz, input[0], temp[10].wwww;
>  2: MAD temp[10].xyz, temp[10].xyz0, const[10].xyz0, const[11].xyz0;
>  3: MOV temp[0], temp[10];
>  4: ADD temp[0].y, const[8], -temp[10];
>  5: TEX temp[1], input[2], 2D[0];
>  6: MOV temp[1].x, temp[1].wwww;
>  7: MAD temp[2], temp[1], const[9].xxxx, -const[9].yyyy;
>  8: MUL temp[2].xy, temp[2], const[1].zzzz;
>  9: MOV temp[2].z, const[9].zzzz;
> 10: DP3 temp[3].x, temp[2], temp[2];
> 11: ADD temp[3].x, const[9].yyyy, -temp[3].xxxx;
> 12: RSQ temp[3].x, |temp[3].xxxx|;
> 13: RCP temp[2].z, temp[3].xxxx;
> 14: TEX temp[4], input[2], 2D[1];
> 15: DP3 temp[3].x, temp[2], input[4];
> 16: DP3 temp[3].y, temp[2], input[5];
> 17: DP3 temp[3].z, temp[2], input[6];
> 18: MUL temp[5], const[1], const[2];
> 19: MAD temp[5], temp[5], temp[3], temp[0];
> 20: DP3 temp[6].w, input[3], input[3];
> 21: RSQ temp[6].w, |temp[6].wwww|;
> 22: MUL temp[6], input[3], temp[6].wwww;
> 23: DP3 temp[2], temp[6], temp[3];
> 24: MUL temp[2], temp[2], const[9].xxxx;
> 25: MAD temp[7], -temp[3], temp[2].xxxx, temp[6];
> 26: DP3_SAT temp[1], temp[7], const[3];
> 27: TEX temp[8], temp[1], 2D[2];
> 28: MUL temp[8], temp[8], const[4];
> 29: MUL temp[8], temp[8], const[5];
> 30: DP3 temp[3], -temp[6], temp[3];
> 31: ADD temp[3], temp[3], const[9].yyyy;
> 32: LG2 temp[11].w, temp[3].xxxx;
> 33: MUL temp[11].w, temp[11].wwww, const[6].xxxx;
> 34: EX2 temp[3], temp[11].wwww;
> 35: RCP_SAT temp[9], temp[3].xxxx;
> 36: TEX temp[3], temp[5], RECT[3];
> 37: MUL temp[3], temp[3], const[7];
> 38: TEX temp[2], temp[7], CUBE[4];
> 39: ADD temp[5], temp[2], -temp[3];
> 40: MAD temp[5], temp[5], temp[9], temp[3];
> 41: ADD output[0], temp[5], temp[8];
> 42: MUL output[0].w, input[1].wwww, temp[4].xxxx;
>Fragment Program: after native rewrite part 2
># Radeon Compiler Program
>  0: RCP temp[10].w, input[0].wwww;
>  1: MUL temp[10].xyz, input[0], temp[10].wwww;
>  2: MAD temp[10].xyz, temp[10].xyz0, const[10].xyz0, const[11].xyz0;
>  3: MOV temp[0], temp[10];
>  4: ADD temp[0].y, const[8], -temp[10];
>  5: TEX temp[1], input[2], 2D[0];
>  6: MOV temp[1].x, temp[1].wwww;
>  7: MAD temp[2], temp[1], const[9].xxxx, -const[9].yyyy;
>  8: MUL temp[2].xy, temp[2], const[1].zzzz;
>  9: MOV temp[2].z, const[9].zzzz;
> 10: DP3 temp[3].x, temp[2], temp[2];
> 11: ADD temp[3].x, const[9].yyyy, -temp[3].xxxx;
> 12: RSQ temp[3].x, |temp[3].xxxx|;
> 13: RCP temp[2].z, temp[3].xxxx;
> 14: TEX temp[4], input[2], 2D[1];
> 15: DP3 temp[3].x, temp[2], input[4];
> 16: DP3 temp[3].y, temp[2], input[5];
> 17: DP3 temp[3].z, temp[2], input[6];
> 18: MUL temp[5], const[1], const[2];
> 19: MAD temp[5], temp[5], temp[3], temp[0];
> 20: DP3 temp[6].w, input[3], input[3];
> 21: RSQ temp[6].w, |temp[6].wwww|;
> 22: MUL temp[6], input[3], temp[6].wwww;
> 23: DP3 temp[2], temp[6], temp[3];
> 24: MUL temp[2], temp[2], const[9].xxxx;
> 25: MAD temp[7], -temp[3], temp[2].xxxx, temp[6];
> 26: DP3_SAT temp[1], temp[7], const[3];
> 27: TEX temp[8], temp[1], 2D[2];
> 28: MUL temp[8], temp[8], const[4];
> 29: MUL temp[8], temp[8], const[5];
> 30: DP3 temp[3], -temp[6], temp[3];
> 31: ADD temp[3], temp[3], const[9].yyyy;
> 32: LG2 temp[11].w, temp[3].xxxx;
> 33: MUL temp[11].w, temp[11].wwww, const[6].xxxx;
> 34: EX2 temp[3], temp[11].wwww;
> 35: RCP_SAT temp[9], temp[3].xxxx;
> 36: MUL temp[12], temp[5], const[12];
> 37: TEX temp[3], temp[12], 2D[3];
> 38: MUL temp[3], temp[3], const[7];
> 39: TEX temp[2], temp[7], CUBE[4];
> 40: ADD temp[5], temp[2], -temp[3];
> 41: MAD temp[5], temp[5], temp[9], temp[3];
> 42: ADD output[0], temp[5], temp[8];
> 43: MUL output[0].w, input[1].wwww, temp[4].xxxx;
>Fragment Program: after deadcode
># Radeon Compiler Program
>  0: RCP temp[10].w, input[0].w___;
>  1: MUL temp[10].xy, input[0].xy__, temp[10].ww__;
>  2: MAD temp[10].xy, temp[10].xy__, const[10].xy__, const[11].xy__;
>  3: MOV temp[0].x, temp[10].x___;
>  4: ADD temp[0].y, const[8]._y__, -temp[10]._y__;
>  5: TEX temp[1].yw, input[2].xy__, 2D[0];
>  6: MOV temp[1].x, temp[1].w___;
>  7: MAD temp[2].xy, temp[1].xy__, const[9].xx__, -const[9].yy__;
>  8: MUL temp[2].xy, temp[2].xy__, const[1].zz__;
>  9: MOV temp[2].z, const[9].__z_;
> 10: DP3 temp[3].x, temp[2].xyz_, temp[2].xyz_;
> 11: ADD temp[3].x, const[9].y___, -temp[3].x___;
> 12: RSQ temp[3].x, |temp[3].x___|;
> 13: RCP temp[2].z, temp[3].x___;
> 14: TEX temp[4].x, input[2].xy__, 2D[1];
> 15: DP3 temp[3].x, temp[2].xyz_, input[4].xyz_;
> 16: DP3 temp[3].y, temp[2].xyz_, input[5].xyz_;
> 17: DP3 temp[3].z, temp[2].xyz_, input[6].xyz_;
> 18: MUL temp[5].xy, const[1].xy__, const[2].xy__;
> 19: MAD temp[5].xy, temp[5].xy__, temp[3].xy__, temp[0].xy__;
> 20: DP3 temp[6].w, input[3].xyz_, input[3].xyz_;
> 21: RSQ temp[6].w, |temp[6].w___|;
> 22: MUL temp[6].xyz, input[3].xyz_, temp[6].www_;
> 23: DP3 temp[2].x, temp[6].xyz_, temp[3].xyz_;
> 24: MUL temp[2].x, temp[2].x___, const[9].x___;
> 25: MAD temp[7].xyz, -temp[3].xyz_, temp[2].xxx_, temp[6].xyz_;
> 26: DP3_SAT temp[1].xy, temp[7].xyz_, const[3].xyz_;
> 27: TEX temp[8].xyz, temp[1].xy__, 2D[2];
> 28: MUL temp[8].xyz, temp[8].xyz_, const[4].xyz_;
> 29: MUL temp[8].xyz, temp[8].xyz_, const[5].xyz_;
> 30: DP3 temp[3].x, -temp[6].xyz_, temp[3].xyz_;
> 31: ADD temp[3].x, temp[3].x___, const[9].y___;
> 32: LG2 temp[11].w, temp[3].x___;
> 33: MUL temp[11].w, temp[11].___w, const[6].___x;
> 34: EX2 temp[3].x, temp[11].w___;
> 35: RCP_SAT temp[9].xyz, temp[3].x___;
> 36: MUL temp[12].xy, temp[5].xy__, const[12].xy__;
> 37: TEX temp[3].xyz, temp[12].xy__, 2D[3];
> 38: MUL temp[3].xyz, temp[3].xyz_, const[7].xyz_;
> 39: TEX temp[2].xyz, temp[7].xyz_, CUBE[4];
> 40: ADD temp[5].xyz, temp[2].xyz_, -temp[3].xyz_;
> 41: MAD temp[5].xyz, temp[5].xyz_, temp[9].xyz_, temp[3].xyz_;
> 42: ADD output[0].xyz, temp[5].xyz_, temp[8].xyz_;
> 43: MUL output[0].w, input[1].___w, temp[4].___x;
>Fragment Program: after dataflow optimize
># Radeon Compiler Program
>  0: RCP temp[10].w, input[0].w___;
>  1: MUL temp[10].xy, input[0].xy__, temp[10].ww__;
>  2: MAD temp[10].xy, temp[10].xy__, const[10].xy__, const[11].xy__;
>  3: MOV temp[0].x, temp[10].x___;
>  4: ADD temp[0].y, const[8]._y__, -temp[10]._y__;
>  5: TEX temp[1].yw, input[2].xy__, 2D[0];
>  6: MOV temp[1].x, temp[1].w___;
>  7: MAD temp[2].xy, temp[1].xy__, const[9].xx__, -none.11__;
>  8: MUL temp[2].xy, temp[2].xy__, const[1].zz__;
>  9: MOV temp[2].z, none.__0_;
> 10: DP3 temp[3].x, temp[2].xyz_, temp[2].xyz_;
> 11: ADD temp[3].x, none.1___, -temp[3].x___;
> 12: RSQ temp[3].x, |temp[3].x___|;
> 13: RCP temp[2].z, temp[3].x___;
> 14: TEX temp[4].x, input[2].xy__, 2D[1];
> 15: DP3 temp[3].x, temp[2].xyz_, input[4].xyz_;
> 16: DP3 temp[3].y, temp[2].xyz_, input[5].xyz_;
> 17: DP3 temp[3].z, temp[2].xyz_, input[6].xyz_;
> 18: MUL temp[5].xy, const[1].xy__, const[2].xy__;
> 19: MAD temp[5].xy, temp[5].xy__, temp[3].xy__, temp[0].xy__;
> 20: DP3 temp[6].w, input[3].xyz_, input[3].xyz_;
> 21: RSQ temp[6].w, |temp[6].w___|;
> 22: MUL temp[6].xyz, input[3].xyz_, temp[6].www_;
> 23: DP3 temp[2].x, temp[6].xyz_, temp[3].xyz_;
> 24: MUL temp[2].x, temp[2].x___, const[9].x___;
> 25: MAD temp[7].xyz, -temp[3].xyz_, temp[2].xxx_, temp[6].xyz_;
> 26: DP3_SAT temp[1].xy, temp[7].xyz_, const[3].xyz_;
> 27: TEX temp[8].xyz, temp[1].xy__, 2D[2];
> 28: MUL temp[8].xyz, temp[8].xyz_, const[4].xyz_;
> 29: MUL temp[8].xyz, temp[8].xyz_, const[5].xyz_;
> 30: DP3 temp[3].x, -temp[6].xyz_, temp[3].xyz_;
> 31: ADD temp[3].x, temp[3].x___, none.1___;
> 32: LG2 temp[11].w, temp[3].x___;
> 33: MUL temp[11].w, temp[11].___w, const[6].___x;
> 34: EX2 temp[3].x, temp[11].w___;
> 35: RCP_SAT temp[9].xyz, temp[3].x___;
> 36: MUL temp[12].xy, temp[5].xy__, const[12].xy__;
> 37: TEX temp[3].xyz, temp[12].xy__, 2D[3];
> 38: MUL temp[3].xyz, temp[3].xyz_, const[7].xyz_;
> 39: TEX temp[2].xyz, temp[7].xyz_, CUBE[4];
> 40: ADD temp[5].xyz, temp[2].xyz_, -temp[3].xyz_;
> 41: MAD temp[5].xyz, temp[5].xyz_, temp[9].xyz_, temp[3].xyz_;
> 42: ADD output[0].xyz, temp[5].xyz_, temp[8].xyz_;
> 43: MUL output[0].w, input[1].___w, temp[4].___x;
>Fragment Program: after dataflow passes
># Radeon Compiler Program
>  0: RCP temp[10].w, input[0].w___;
>  1: MUL temp[10].xy, input[0].xy__, temp[10].ww__;
>  2: MAD temp[10].xy, temp[10].xy__, const[10].xy__, const[11].xy__;
>  3: MOV temp[0].x, temp[10].x___;
>  4: ADD temp[0].y, const[8]._y__, -temp[10]._y__;
>  5: TEX temp[1].yw, input[2].xy__, 2D[0];
>  6: MOV temp[1].x, temp[1].w___;
>  7: MAD temp[2].xy, temp[1].xy__, const[9].xx__, -none.11__;
>  8: MUL temp[2].xy, temp[2].xy__, const[1].zz__;
>  9: MOV temp[2].z, none.__0_;
> 10: DP3 temp[3].x, temp[2].xyz_, temp[2].xyz_;
> 11: ADD temp[3].x, none.1___, -temp[3].x___;
> 12: RSQ temp[3].x, |temp[3].x___|;
> 13: RCP temp[2].z, temp[3].x___;
> 14: TEX temp[4].x, input[2].xy__, 2D[1];
> 15: DP3 temp[3].x, temp[2].xyz_, input[4].xyz_;
> 16: DP3 temp[3].y, temp[2].xyz_, input[5].xyz_;
> 17: DP3 temp[3].z, temp[2].xyz_, input[6].xyz_;
> 18: MUL temp[5].xy, const[1].xy__, const[2].xy__;
> 19: MAD temp[5].xy, temp[5].xy__, temp[3].xy__, temp[0].xy__;
> 20: DP3 temp[6].w, input[3].xyz_, input[3].xyz_;
> 21: RSQ temp[6].w, |temp[6].w___|;
> 22: MUL temp[6].xyz, input[3].xyz_, temp[6].www_;
> 23: DP3 temp[2].x, temp[6].xyz_, temp[3].xyz_;
> 24: MUL temp[2].x, temp[2].x___, const[9].x___;
> 25: MAD temp[7].xyz, -temp[3].xyz_, temp[2].xxx_, temp[6].xyz_;
> 26: DP3_SAT temp[1].xy, temp[7].xyz_, const[3].xyz_;
> 27: TEX temp[8].xyz, temp[1].xy__, 2D[2];
> 28: MUL temp[8].xyz, temp[8].xyz_, const[4].xyz_;
> 29: MUL temp[8].xyz, temp[8].xyz_, const[5].xyz_;
> 30: DP3 temp[3].x, -temp[6].xyz_, temp[3].xyz_;
> 31: ADD temp[3].x, temp[3].x___, none.1___;
> 32: LG2 temp[11].w, temp[3].x___;
> 33: MUL temp[11].w, temp[11].___w, const[6].___x;
> 34: EX2 temp[3].x, temp[11].w___;
> 35: RCP_SAT temp[9].xyz, temp[3].x___;
> 36: MUL temp[12].xy, temp[5].xy__, const[12].xy__;
> 37: TEX temp[3].xyz, temp[12].xy__, 2D[3];
> 38: MUL temp[3].xyz, temp[3].xyz_, const[7].xyz_;
> 39: TEX temp[2].xyz, temp[7].xyz_, CUBE[4];
> 40: ADD temp[5].xyz, temp[2].xyz_, -temp[3].xyz_;
> 41: MAD temp[5].xyz, temp[5].xyz_, temp[9].xyz_, temp[3].xyz_;
> 42: ADD output[0].xyz, temp[5].xyz_, temp[8].xyz_;
> 43: MUL output[0].w, input[1].___w, temp[4].___x;
>Fragment Program: after pair translate
># Radeon Compiler Program
>  0: src0.w = input[0]
>     RCP temp[10].w, src0.w
>  1: src0.xyz = input[0], src0.w = temp[10]
>     MAD temp[10].xy, src0.xy_, src0.ww_, src0.000
>  2: src0.xyz = temp[10], src1.xyz = const[10], src2.xyz = const[11]
>     MAD temp[10].xy, src0.xy_, src1.xy_, src2.xy_
>  3: src0.xyz = temp[10]
>     MAD temp[0].x, src0.x__, src0.111, src0.000
>  4: src0.xyz = const[8], src1.xyz = temp[10]
>     MAD temp[0].y, src0._y_, src0.111, -src1._y_
>  5: TEX temp[1].yw, input[2].xy__, 2D[0];
>  6: src0.w = temp[1]
>     MAD temp[1].x, src0.w__, src0.111, src0.000
>  7: src0.xyz = temp[1], src1.xyz = const[9]
>     MAD temp[2].xy, src0.xy_, src1.xx_, -src0.11_
>  8: src0.xyz = temp[2], src1.xyz = const[1]
>     MAD temp[2].xy, src0.xy_, src1.zz_, src0.000
>  9: 
>     MAD temp[2].z, src0.__0, src0.111, src0.000
> 10: src0.xyz = temp[2]
>     DP3 temp[3].x, src0.xyz, src0.xyz
> 11: src0.xyz = temp[3]
>     MAD temp[3].x, src0.1__, src0.111, -src0.x__
> 12: src0.xyz = temp[3]
>     REPL_ALPHA temp[3].x
>     RSQ, |src0.x|
> 13: src0.xyz = temp[3]
>     REPL_ALPHA temp[2].z
>     RCP, src0.x
> 14: TEX temp[4].x, input[2].xy__, 2D[1];
> 15: src0.xyz = temp[2], src1.xyz = input[4]
>     DP3 temp[3].x, src0.xyz, src1.xyz
> 16: src0.xyz = temp[2], src1.xyz = input[5]
>     DP3 temp[3].y, src0.xyz, src1.xyz
> 17: src0.xyz = temp[2], src1.xyz = input[6]
>     DP3 temp[3].z, src0.xyz, src1.xyz
> 18: src0.xyz = const[1], src1.xyz = const[2]
>     MAD temp[5].xy, src0.xy_, src1.xy_, src0.000
> 19: src0.xyz = temp[5], src1.xyz = temp[3], src2.xyz = temp[0]
>     MAD temp[5].xy, src0.xy_, src1.xy_, src2.xy_
> 20: src0.xyz = input[3]
>     DP3, src0.xyz, src0.xyz
>     DP3 temp[6].w, src0._, src0._
> 21: src0.w = temp[6]
>     RSQ temp[6].w, |src0.w|
> 22: src0.xyz = input[3], src0.w = temp[6]
>     MAD temp[6].xyz, src0.xyz, src0.www, src0.000
> 23: src0.xyz = temp[6], src1.xyz = temp[3]
>     DP3 temp[2].x, src0.xyz, src1.xyz
> 24: src0.xyz = temp[2], src1.xyz = const[9]
>     MAD temp[2].x, src0.x__, src1.x__, src0.000
> 25: src0.xyz = temp[3], src1.xyz = temp[2], src2.xyz = temp[6]
>     MAD temp[7].xyz, -src0.xyz, src1.xxx, src2.xyz
> 26: src0.xyz = temp[7], src1.xyz = const[3]
>     DP3_SAT temp[1].xy, src0.xyz, src1.xyz
> 27: TEX temp[8].xyz, temp[1].xy__, 2D[2];
> 28: src0.xyz = temp[8], src1.xyz = const[4]
>     MAD temp[8].xyz, src0.xyz, src1.xyz, src0.000
> 29: src0.xyz = temp[8], src1.xyz = const[5]
>     MAD temp[8].xyz, src0.xyz, src1.xyz, src0.000
> 30: src0.xyz = temp[6], src1.xyz = temp[3]
>     DP3 temp[3].x, -src0.xyz, src1.xyz
> 31: src0.xyz = temp[3]
>     MAD temp[3].x, src0.x__, src0.111, src0.1__
> 32: src0.xyz = temp[3]
>     LG2 temp[11].w, src0.x
> 33: src0.xyz = const[6], src0.w = temp[11]
>     MAD temp[11].w, src0.w, src0.x, src0.0
> 34: src0.w = temp[11]
>     REPL_ALPHA temp[3].x
>     EX2, src0.w
> 35: src0.xyz = temp[3]
>     REPL_ALPHA_SAT temp[9].xyz
>     RCP_SAT, src0.x
> 36: src0.xyz = temp[5], src1.xyz = const[12]
>     MAD temp[12].xy, src0.xy_, src1.xy_, src0.000
> 37: TEX temp[3].xyz, temp[12].xy__, 2D[3];
> 38: src0.xyz = temp[3], src1.xyz = const[7]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src0.000
> 39: TEX temp[2].xyz, temp[7].xyz_, CUBE[4];
> 40: src0.xyz = temp[2], src1.xyz = temp[3]
>     MAD temp[5].xyz, src0.xyz, src0.111, -src1.xyz
> 41: src0.xyz = temp[5], src1.xyz = temp[9], src2.xyz = temp[3]
>     MAD temp[5].xyz, src0.xyz, src1.xyz, src2.xyz
> 42: src0.xyz = temp[5], src1.xyz = temp[8]
>     MAD color[0].xyz, src0.xyz, src0.111, src1.xyz
> 43: src0.xyz = temp[4], src0.w = input[1]
>     MAD color[0].w, src0.w, src0.x, src0.0
>Fragment Program: after pair scheduling
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[4].x, input[2].xy__, 2D[1];
>  2: TEX temp[1].yw, input[2].xy__, 2D[0];
>  3: src0.xyz = input[3]
>     DP3, src0.xyz, src0.xyz
>     DP3 temp[6].w, src0._, src0._
>  4: src0.w = temp[1], src1.w = temp[6]
>     MAD temp[1].x, src0.w__, src0.111, src0.000
>     RSQ temp[6].w, |src1.w|
>  5: src0.xyz = input[3], src0.w = temp[6], src1.xyz = temp[4], src1.w = input[1]
>     MAD temp[6].xyz, src0.xyz, src0.www, src0.000
>     MAD color[0].w, src1.w, src1.x, src0.0
>  6: src0.xyz = temp[1], src0.w = input[0], src1.xyz = const[9]
>     MAD temp[2].xy, src0.xy_, src1.xx_, -src0.11_
>     RCP temp[10].w, src0.w
>  7: src0.xyz = input[0], src0.w = temp[10]
>     MAD temp[10].xy, src0.xy_, src0.ww_, src0.000
>  8: src0.xyz = temp[10], src1.xyz = const[10], src2.xyz = const[11]
>     MAD temp[10].xy, src0.xy_, src1.xy_, src2.xy_
>  9: src0.xyz = const[8], src1.xyz = temp[10]
>     MAD temp[0].y, src0._y_, src0.111, -src1._y_
> 10: src0.xyz = temp[10]
>     MAD temp[0].x, src0.x__, src0.111, src0.000
> 11: src0.xyz = temp[2], src1.xyz = const[1]
>     MAD temp[2].xy, src0.xy_, src1.zz_, src0.000
> 12: src0.xyz = const[1], src1.xyz = const[2]
>     MAD temp[5].xy, src0.xy_, src1.xy_, src0.000
> 13: 
>     MAD temp[2].z, src0.__0, src0.111, src0.000
> 14: src0.xyz = temp[2]
>     DP3 temp[3].x, src0.xyz, src0.xyz
> 15: src0.xyz = temp[3]
>     MAD temp[3].x, src0.1__, src0.111, -src0.x__
> 16: src0.xyz = temp[3]
>     REPL_ALPHA temp[3].x
>     RSQ, |src0.x|
> 17: src0.xyz = temp[3]
>     REPL_ALPHA temp[2].z
>     RCP, src0.x
> 18: src0.xyz = temp[2], src1.xyz = input[4]
>     DP3 temp[3].x, src0.xyz, src1.xyz
> 19: src0.xyz = temp[2], src1.xyz = input[5]
>     DP3 temp[3].y, src0.xyz, src1.xyz
> 20: src0.xyz = temp[5], src1.xyz = temp[3], src2.xyz = temp[0]
>     MAD temp[5].xy, src0.xy_, src1.xy_, src2.xy_
> 21: src0.xyz = temp[5], src1.xyz = const[12]
>     MAD temp[12].xy, src0.xy_, src1.xy_, src0.000
> 22: src0.xyz = temp[2], src1.xyz = input[6]
>     DP3 temp[3].z, src0.xyz, src1.xyz
> 23: src0.xyz = temp[6], src1.xyz = temp[3]
>     DP3 temp[2].x, src0.xyz, src1.xyz
> 24: src0.xyz = temp[2], src1.xyz = const[9]
>     MAD temp[2].x, src0.x__, src1.x__, src0.000
> 25: src0.xyz = temp[3], src1.xyz = temp[2], src2.xyz = temp[6]
>     MAD temp[7].xyz, -src0.xyz, src1.xxx, src2.xyz
> 26: src0.xyz = temp[7], src1.xyz = const[3]
>     DP3_SAT temp[1].xy, src0.xyz, src1.xyz
> 27: src0.xyz = temp[6], src1.xyz = temp[3]
>     DP3 temp[3].x, -src0.xyz, src1.xyz
> 28: src0.xyz = temp[3]
>     MAD temp[3].x, src0.x__, src0.111, src0.1__
> 29: src0.xyz = temp[3]
>     LG2 temp[11].w, src0.x
> 30: src0.xyz = const[6], src0.w = temp[11]
>     MAD temp[11].w, src0.w, src0.x, src0.0
> 31: src0.w = temp[11]
>     REPL_ALPHA temp[3].x
>     EX2, src0.w
> 32: src0.xyz = temp[3]
>     REPL_ALPHA_SAT temp[9].xyz
>     RCP_SAT, src0.x
> 33: BEGIN_TEX;
> 34: TEX temp[3].xyz, temp[12].xy__, 2D[3];
> 35: TEX temp[8].xyz, temp[1].xy__, 2D[2];
> 36: TEX temp[2].xyz, temp[7].xyz_, CUBE[4];
> 37: src0.xyz = temp[8], src1.xyz = const[4]
>     MAD temp[8].xyz, src0.xyz, src1.xyz, src0.000
> 38: src0.xyz = temp[8], src1.xyz = const[5]
>     MAD temp[8].xyz, src0.xyz, src1.xyz, src0.000
> 39: src0.xyz = temp[3], src1.xyz = const[7]
>     MAD temp[3].xyz, src0.xyz, src1.xyz, src0.000
> 40: src0.xyz = temp[2], src1.xyz = temp[3]
>     MAD temp[5].xyz, src0.xyz, src0.111, -src1.xyz
> 41: src0.xyz = temp[5], src1.xyz = temp[9], src2.xyz = temp[3]
>     MAD temp[5].xyz, src0.xyz, src1.xyz, src2.xyz
> 42: src0.xyz = temp[5], src1.xyz = temp[8]
>     MAD color[0].xyz, src0.xyz, src0.111, src1.xyz
>Fragment Program: after register allocation
># Radeon Compiler Program
>  0: BEGIN_TEX;
>  1: TEX temp[7].x, temp[1].xy__, 2D[1];
>  2: TEX temp[1].yw, temp[1].xy__, 2D[0];
>  3: src0.xyz = temp[2]
>     DP3, src0.xyz, src0.xyz
>     DP3 temp[8].w, src0._, src0._
>  4: src0.w = temp[1], src1.w = temp[8]
>     MAD temp[1].x, src0.w__, src0.111, src0.000
>     RSQ temp[8].w, |src1.w|
>  5: src0.xyz = temp[2], src0.w = temp[8], src1.xyz = temp[7], src1.w = temp[0]
>     MAD temp[8].xyz, src0.xyz, src0.www, src0.000
>     MAD color[0].w, src1.w, src1.x, src0.0
>  6: src0.xyz = temp[1], src0.w = temp[6], src1.xyz = const[9]
>     MAD temp[2].xy, src0.xy_, src1.xx_, -src0.11_
>     RCP temp[7].w, src0.w
>  7: src0.xyz = temp[6], src0.w = temp[7]
>     MAD temp[7].xy, src0.xy_, src0.ww_, src0.000
>  8: src0.xyz = temp[7], src1.xyz = const[10], src2.xyz = const[11]
>     MAD temp[7].xy, src0.xy_, src1.xy_, src2.xy_
>  9: src0.xyz = const[8], src1.xyz = temp[7]
>     MAD temp[0].y, src0._y_, src0.111, -src1._y_
> 10: src0.xyz = temp[7]
>     MAD temp[0].x, src0.x__, src0.111, src0.000
> 11: src0.xyz = temp[2], src1.xyz = const[1]
>     MAD temp[2].xy, src0.xy_, src1.zz_, src0.000
> 12: src0.xyz = const[1], src1.xyz = const[2]
>     MAD temp[7].xy, src0.xy_, src1.xy_, src0.000
> 13: 
>     MAD temp[2].z, src0.__0, src0.111, src0.000
> 14: src0.xyz = temp[2]
>     DP3 temp[6].x, src0.xyz, src0.xyz
> 15: src0.xyz = temp[6]
>     MAD temp[6].x, src0.1__, src0.111, -src0.x__
> 16: src0.xyz = temp[6]
>     REPL_ALPHA temp[6].x
>     RSQ, |src0.x|
> 17: src0.xyz = temp[6]
>     REPL_ALPHA temp[2].z
>     RCP, src0.x
> 18: src0.xyz = temp[2], src1.xyz = temp[3]
>     DP3 temp[6].x, src0.xyz, src1.xyz
> 19: src0.xyz = temp[2], src1.xyz = temp[4]
>     DP3 temp[6].y, src0.xyz, src1.xyz
> 20: src0.xyz = temp[7], src1.xyz = temp[6], src2.xyz = temp[0]
>     MAD temp[7].xy, src0.xy_, src1.xy_, src2.xy_
> 21: src0.xyz = temp[7], src1.xyz = const[12]
>     MAD temp[4].xy, src0.xy_, src1.xy_, src0.000
> 22: src0.xyz = temp[2], src1.xyz = temp[5]
>     DP3 temp[6].z, src0.xyz, src1.xyz
> 23: src0.xyz = temp[8], src1.xyz = temp[6]
>     DP3 temp[2].x, src0.xyz, src1.xyz
> 24: src0.xyz = temp[2], src1.xyz = const[9]
>     MAD temp[2].x, src0.x__, src1.x__, src0.000
> 25: src0.xyz = temp[6], src1.xyz = temp[2], src2.xyz = temp[8]
>     MAD temp[0].xyz, -src0.xyz, src1.xxx, src2.xyz
> 26: src0.xyz = temp[0], src1.xyz = const[3]
>     DP3_SAT temp[1].xy, src0.xyz, src1.xyz
> 27: src0.xyz = temp[8], src1.xyz = temp[6]
>     DP3 temp[6].x, -src0.xyz, src1.xyz
> 28: src0.xyz = temp[6]
>     MAD temp[6].x, src0.x__, src0.111, src0.1__
> 29: src0.xyz = temp[6]
>     LG2 temp[3].w, src0.x
> 30: src0.xyz = const[6], src0.w = temp[3]
>     MAD temp[3].w, src0.w, src0.x, src0.0
> 31: src0.w = temp[3]
>     REPL_ALPHA temp[6].x
>     EX2, src0.w
> 32: src0.xyz = temp[6]
>     REPL_ALPHA_SAT temp[3].xyz
>     RCP_SAT, src0.x
> 33: BEGIN_TEX;
> 34: TEX temp[6].xyz, temp[4].xy__, 2D[3];
> 35: TEX temp[1].xyz, temp[1].xy__, 2D[2];
> 36: TEX temp[2].xyz, temp[0].xyz_, CUBE[4];
> 37: src0.xyz = temp[1], src1.xyz = const[4]
>     MAD temp[1].xyz, src0.xyz, src1.xyz, src0.000
> 38: src0.xyz = temp[1], src1.xyz = const[5]
>     MAD temp[1].xyz, src0.xyz, src1.xyz, src0.000
> 39: src0.xyz = temp[6], src1.xyz = const[7]
>     MAD temp[6].xyz, src0.xyz, src1.xyz, src0.000
> 40: src0.xyz = temp[2], src1.xyz = temp[6]
>     MAD temp[7].xyz, src0.xyz, src0.111, -src1.xyz
> 41: src0.xyz = temp[7], src1.xyz = temp[3], src2.xyz = temp[6]
>     MAD temp[7].xyz, src0.xyz, src1.xyz, src2.xyz
> 42: src0.xyz = temp[7], src1.xyz = temp[1]
>     MAD color[0].xyz, src0.xyz, src0.111, src1.xyz
>pc=54*************************************
>Hardware program
>----------------
>NODE 0: alu_offset: 0, tex_offset: 0, alu_end: 29, tex_end: 1  (code_addr: 00420740)
>  TEX:
>    TEX t7, t1, texture[1]   (000089c1)
>    TEX t1, t1, texture[0]   (00008041)
>  0: xyz:  t2  t0  t0 ->                      (00000002)
>       w:  t0  t0  t0 -> t8.w                 (00a00000)
>     xyz:   t2.xyz   t2.xyz   t2.xxx    op: 00804000
>       w:      1.0      1.0     t2.x    op: 00800891
>  1: xyz:  t0  t0  t0 -> t1.x                 (00840000)
>       w:  t1  t8  t0 -> t8.w                 (00a00201)
>     xyz:   t1.www      1.0      0.0    op: 00050a8c
>       w:   |t8.w|     t0.x     t0.x    op: 0580004a
>  2: xyz:  t2  t7  t0 -> t8.xyz               (03a001c2)
>       w:  t8  t0  t0 -> o0.w                 (01000008)
>     xyz:   t2.xyz   t8.www      0.0    op: 00050600
>       w:     t0.w     t7.x      0.0    op: 0004018a
>  3: xyz:  t1  c9  t0 -> t2.xy                (01880a41)
>       w:  t6  t0  t0 -> t7.w                 (009c0006)
>     xyz:   t1.xyz   c9.xxx     -1.0    op: 000d4280
>       w:     t6.w     t1.x     t1.x    op: 05000009
>  4: xyz:  t6  t0  t0 -> t7.xy                (019c0006)
>       w:  t7  t0  t0 ->                      (00000007)
>     xyz:   t6.xyz   t7.www      0.0    op: 00050600
>       w:     t6.x     t6.x     t6.x    op: 00000000
>  5: xyz:  t7 c10 c11 -> t7.xy                (019eba87)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t7.xyz  c10.xyz  c11.xyz    op: 00020200
>       w:     t7.x     t7.x     t7.x    op: 00000000
>  6: xyz:  c8  t7  t0 -> t0.y                 (010001e8)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   c8.xyz      1.0  -t7.xyz    op: 00090a80
>       w:     c8.x     c8.x     c8.x    op: 00000000
>  7: xyz:  t7  t0  t0 -> t0.x                 (00800007)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t7.xyz      1.0      0.0    op: 00050a80
>       w:     t7.x     t7.x     t7.x    op: 00000000
>  8: xyz:  t2  c1  t0 -> t2.xy                (01880842)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz   c1.zzz      0.0    op: 00050380
>       w:     t2.x     t2.x     t2.x    op: 00000000
>  9: xyz:  c1  c2  t0 -> t7.xy                (019c08a1)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   c1.xyz   c2.xyz      0.0    op: 00050200
>       w:     c1.x     c1.x     c1.x    op: 00000000
> 10: xyz:  t0  t0  t0 -> t2.z                 (02080000)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:      0.0      1.0      0.0    op: 00050a94
>       w:     t0.x     t0.x     t0.x    op: 00000000
> 11: xyz:  t2  t0  t0 -> t6.x                 (00980002)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz   t2.xyz   t2.xxx    op: 00804000
>       w:     t2.x     t2.x     t2.x    op: 00000000
> 12: xyz:  t6  t0  t0 -> t6.x                 (00980006)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:      1.0      1.0  -t6.xyz    op: 00080a95
>       w:     t6.x     t6.x     t6.x    op: 00000000
> 13: xyz:  t6  t0  t0 -> t6.x                 (00980006)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t6.xxx   t6.xxx   t6.xxx    op: 05004081
>       w:   |t6.x|     t6.x     t6.x    op: 05800040
> 14: xyz:  t6  t0  t0 -> t2.z                 (02080006)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t6.xxx   t6.xxx   t6.xxx    op: 05004081
>       w:     t6.x     t6.x     t6.x    op: 05000000
> 15: xyz:  t2  t3  t0 -> t6.x                 (009800c2)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz   t3.xyz   t2.xxx    op: 00804200
>       w:     t2.x     t2.x     t2.x    op: 00000000
> 16: xyz:  t2  t4  t0 -> t6.y                 (01180102)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz   t4.xyz   t2.xxx    op: 00804200
>       w:     t2.x     t2.x     t2.x    op: 00000000
> 17: xyz:  t7  t6  t0 -> t7.xy                (019c0187)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t7.xyz   t6.xyz   t0.xyz    op: 00020200
>       w:     t7.x     t7.x     t7.x    op: 00000000
> 18: xyz:  t7 c12  t0 -> t4.xy                (01900b07)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t7.xyz  c12.xyz      0.0    op: 00050200
>       w:     t7.x     t7.x     t7.x    op: 00000000
> 19: xyz:  t2  t5  t0 -> t6.z                 (02180142)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz   t5.xyz   t2.xxx    op: 00804200
>       w:     t2.x     t2.x     t2.x    op: 00000000
> 20: xyz:  t8  t6  t0 -> t2.x                 (00880188)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t8.xyz   t6.xyz   t8.xxx    op: 00804200
>       w:     t8.x     t8.x     t8.x    op: 00000000
> 21: xyz:  t2  c9  t0 -> t2.x                 (00880a42)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz   c9.xyz      0.0    op: 00050200
>       w:     t2.x     t2.x     t2.x    op: 00000000
> 22: xyz:  t6  t2  t8 -> t0.xyz               (03808086)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:  -t6.xyz   t2.xxx   t8.xyz    op: 000202a0
>       w:     t6.x     t6.x     t6.x    op: 00000000
> 23: xyz:  t0  c3  t0 -> t1.xy                (018408c0)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t0.xyz   c3.xyz   t0.xxx    op: 40804200
>       w:     t0.x     t0.x     t0.x    op: 00000000
> 24: xyz:  t8  t6  t0 -> t6.x                 (00980188)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:  -t8.xyz   t6.xyz   t8.xxx    op: 00804220
>       w:     t8.x     t8.x     t8.x    op: 00000000
> 25: xyz:  t6  t0  t0 -> t6.x                 (00980006)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t6.xyz      1.0      1.0    op: 00054a80
>       w:     t6.x     t6.x     t6.x    op: 00000000
> 26: xyz:  t6  t0  t0 ->                      (00000006)
>       w:  t0  t0  t0 -> t3.w                 (008c0000)
>     xyz:   t6.xxx   t6.xxx   t6.xxx    op: 00004081
>       w:     t6.x     t6.x     t6.x    op: 04800000
> 27: xyz:  c6  t0  t0 ->                      (00000026)
>       w:  t3  t0  t0 -> t3.w                 (008c0003)
>     xyz:   c6.xxx   c6.xxx   c6.xxx    op: 00004081
>       w:     t3.w     c6.x      0.0    op: 00040009
> 28: xyz:  t0  t0  t0 -> t6.x                 (00980000)
>       w:  t3  t0  t0 ->                      (00000003)
>     xyz:   t0.xxx   t0.xxx   t0.xxx    op: 05004081
>       w:     t3.w     t0.x     t0.x    op: 04000009
> 29: xyz:  t6  t0  t0 -> t3.xyz               (038c0006)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t6.xxx   t6.xxx   t6.xxx    op: 45004081
>       w:     t6.x     t6.x     t6.x    op: 45000000
>NODE 1: alu_offset: 30, tex_offset: 2, alu_end: 5, tex_end: 2  (code_addr: 0044215e)
>  TEX:
>    TEX t6, t4, texture[3]   (00009984)
>    TEX t1, t1, texture[2]   (00009041)
>    TEX t2, t0, texture[4]   (0000a080)
> 30: xyz:  t1  c4  t0 -> t1.xyz               (03840901)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   c4.xyz      0.0    op: 00050200
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 31: xyz:  t1  c5  t0 -> t1.xyz               (03840941)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t1.xyz   c5.xyz      0.0    op: 00050200
>       w:     t1.x     t1.x     t1.x    op: 00000000
> 32: xyz:  t6  c7  t0 -> t6.xyz               (039809c6)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t6.xyz   c7.xyz      0.0    op: 00050200
>       w:     t6.x     t6.x     t6.x    op: 00000000
> 33: xyz:  t2  t6  t0 -> t7.xyz               (039c0182)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t2.xyz      1.0  -t6.xyz    op: 00090a80
>       w:     t2.x     t2.x     t2.x    op: 00000000
> 34: xyz:  t7  t3  t6 -> t7.xyz               (039c60c7)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t7.xyz   t3.xyz   t6.xyz    op: 00020200
>       w:     t7.x     t7.x     t7.x    op: 00000000
> 35: xyz:  t7  t1  t0 -> o0.xyz               (1c000047)
>       w:  t0  t0  t0 ->                      (00000000)
>     xyz:   t7.xyz      1.0   t1.xyz    op: 00010a80
>       w:     t7.x     t7.x     t7.x    op: 00000000

Actions: View

Attachments on bug 28624: 36379

Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.