4 TGSI, Tungsten Graphics Shader Infrastructure, is an intermediate language
5 for describing shaders. Since Gallium is inherently shaderful, shaders are
6 an important part of the API. TGSI is the only intermediate representation
12 All TGSI instructions, known as *opcodes*, operate on arbitrary-precision
13 floating-point four-component vectors. An opcode may have up to one
14 destination register, known as *dst*, and between zero and three source
15 registers, called *src0* through *src2*, or simply *src* if there is only
18 Some instructions, like :opcode:`I2F`, permit re-interpretation of vector
19 components as integers. Other instructions permit using registers as
20 two-component vectors with double precision; see :ref:`doubleopcodes`.
22 When an instruction has a scalar result, the result is usually copied into
23 each of the components of *dst*. When this happens, the result is said to be
24 *replicated* to *dst*. :opcode:`RCP` is one such instruction.
29 TGSI supports modifiers on inputs (as well as saturate and precise modifier
32 For arithmetic instruction having a precise modifier certain optimizations
33 which may alter the result are disallowed. Example: *add(mul(a,b),c)* can't be
34 optimized to TGSI_OPCODE_MAD, because some hardware only supports the fused
37 For inputs which have a floating point type, both absolute value and
38 negation modifiers are supported (with absolute value being applied
39 first). The only source of TGSI_OPCODE_MOV and the second and third
40 sources of TGSI_OPCODE_UCMP are considered to have float type for
43 For inputs which have signed or unsigned type only the negate modifier is
50 ^^^^^^^^^^^^^^^^^^^^^^^^^
52 These opcodes are guaranteed to be available regardless of the driver being
55 .. opcode:: ARL - Address Register Load
59 dst.x = (int) \lfloor src.x\rfloor
61 dst.y = (int) \lfloor src.y\rfloor
63 dst.z = (int) \lfloor src.z\rfloor
65 dst.w = (int) \lfloor src.w\rfloor
68 .. opcode:: MOV - Move
81 .. opcode:: LIT - Light Coefficients
86 dst.y &= max(src.x, 0) \\
87 dst.z &= (src.x > 0) ? max(src.y, 0)^{clamp(src.w, -128, 128))} : 0 \\
91 .. opcode:: RCP - Reciprocal
93 This instruction replicates its result.
100 .. opcode:: RSQ - Reciprocal Square Root
102 This instruction replicates its result. The results are undefined for src <= 0.
106 dst = \frac{1}{\sqrt{src.x}}
109 .. opcode:: SQRT - Square Root
111 This instruction replicates its result. The results are undefined for src < 0.
118 .. opcode:: EXP - Approximate Exponential Base 2
122 dst.x &= 2^{\lfloor src.x\rfloor} \\
123 dst.y &= src.x - \lfloor src.x\rfloor \\
124 dst.z &= 2^{src.x} \\
128 .. opcode:: LOG - Approximate Logarithm Base 2
132 dst.x &= \lfloor\log_2{|src.x|}\rfloor \\
133 dst.y &= \frac{|src.x|}{2^{\lfloor\log_2{|src.x|}\rfloor}} \\
134 dst.z &= \log_2{|src.x|} \\
138 .. opcode:: MUL - Multiply
142 dst.x = src0.x \times src1.x
144 dst.y = src0.y \times src1.y
146 dst.z = src0.z \times src1.z
148 dst.w = src0.w \times src1.w
151 .. opcode:: ADD - Add
155 dst.x = src0.x + src1.x
157 dst.y = src0.y + src1.y
159 dst.z = src0.z + src1.z
161 dst.w = src0.w + src1.w
164 .. opcode:: DP3 - 3-component Dot Product
166 This instruction replicates its result.
170 dst = src0.x \times src1.x + src0.y \times src1.y + src0.z \times src1.z
173 .. opcode:: DP4 - 4-component Dot Product
175 This instruction replicates its result.
179 dst = src0.x \times src1.x + src0.y \times src1.y + src0.z \times src1.z + src0.w \times src1.w
182 .. opcode:: DST - Distance Vector
187 dst.y &= src0.y \times src1.y\\
192 .. opcode:: MIN - Minimum
196 dst.x = min(src0.x, src1.x)
198 dst.y = min(src0.y, src1.y)
200 dst.z = min(src0.z, src1.z)
202 dst.w = min(src0.w, src1.w)
205 .. opcode:: MAX - Maximum
209 dst.x = max(src0.x, src1.x)
211 dst.y = max(src0.y, src1.y)
213 dst.z = max(src0.z, src1.z)
215 dst.w = max(src0.w, src1.w)
218 .. opcode:: SLT - Set On Less Than
222 dst.x = (src0.x < src1.x) ? 1.0F : 0.0F
224 dst.y = (src0.y < src1.y) ? 1.0F : 0.0F
226 dst.z = (src0.z < src1.z) ? 1.0F : 0.0F
228 dst.w = (src0.w < src1.w) ? 1.0F : 0.0F
231 .. opcode:: SGE - Set On Greater Equal Than
235 dst.x = (src0.x >= src1.x) ? 1.0F : 0.0F
237 dst.y = (src0.y >= src1.y) ? 1.0F : 0.0F
239 dst.z = (src0.z >= src1.z) ? 1.0F : 0.0F
241 dst.w = (src0.w >= src1.w) ? 1.0F : 0.0F
244 .. opcode:: MAD - Multiply And Add
246 Perform a * b + c. The implementation is free to decide whether there is an
247 intermediate rounding step or not.
251 dst.x = src0.x \times src1.x + src2.x
253 dst.y = src0.y \times src1.y + src2.y
255 dst.z = src0.z \times src1.z + src2.z
257 dst.w = src0.w \times src1.w + src2.w
260 .. opcode:: LRP - Linear Interpolate
264 dst.x = src0.x \times src1.x + (1 - src0.x) \times src2.x
266 dst.y = src0.y \times src1.y + (1 - src0.y) \times src2.y
268 dst.z = src0.z \times src1.z + (1 - src0.z) \times src2.z
270 dst.w = src0.w \times src1.w + (1 - src0.w) \times src2.w
273 .. opcode:: FMA - Fused Multiply-Add
275 Perform a * b + c with no intermediate rounding step.
279 dst.x = src0.x \times src1.x + src2.x
281 dst.y = src0.y \times src1.y + src2.y
283 dst.z = src0.z \times src1.z + src2.z
285 dst.w = src0.w \times src1.w + src2.w
288 .. opcode:: FRC - Fraction
292 dst.x = src.x - \lfloor src.x\rfloor
294 dst.y = src.y - \lfloor src.y\rfloor
296 dst.z = src.z - \lfloor src.z\rfloor
298 dst.w = src.w - \lfloor src.w\rfloor
301 .. opcode:: FLR - Floor
305 dst.x = \lfloor src.x\rfloor
307 dst.y = \lfloor src.y\rfloor
309 dst.z = \lfloor src.z\rfloor
311 dst.w = \lfloor src.w\rfloor
314 .. opcode:: ROUND - Round
327 .. opcode:: EX2 - Exponential Base 2
329 This instruction replicates its result.
336 .. opcode:: LG2 - Logarithm Base 2
338 This instruction replicates its result.
345 .. opcode:: POW - Power
347 This instruction replicates its result.
351 dst = src0.x^{src1.x}
353 .. opcode:: XPD - Cross Product
357 dst.x = src0.y \times src1.z - src1.y \times src0.z
359 dst.y = src0.z \times src1.x - src1.z \times src0.x
361 dst.z = src0.x \times src1.y - src1.x \times src0.y
366 .. opcode:: COS - Cosine
368 This instruction replicates its result.
375 .. opcode:: DDX, DDX_FINE - Derivative Relative To X
377 The fine variant is only used when ``PIPE_CAP_TGSI_FS_FINE_DERIVATIVE`` is
378 advertised. When it is, the fine version guarantees one derivative per row
379 while DDX is allowed to be the same for the entire 2x2 quad.
383 dst.x = partialx(src.x)
385 dst.y = partialx(src.y)
387 dst.z = partialx(src.z)
389 dst.w = partialx(src.w)
392 .. opcode:: DDY, DDY_FINE - Derivative Relative To Y
394 The fine variant is only used when ``PIPE_CAP_TGSI_FS_FINE_DERIVATIVE`` is
395 advertised. When it is, the fine version guarantees one derivative per column
396 while DDY is allowed to be the same for the entire 2x2 quad.
400 dst.x = partialy(src.x)
402 dst.y = partialy(src.y)
404 dst.z = partialy(src.z)
406 dst.w = partialy(src.w)
409 .. opcode:: PK2H - Pack Two 16-bit Floats
411 This instruction replicates its result.
415 dst = f32\_to\_f16(src.x) | f32\_to\_f16(src.y) << 16
418 .. opcode:: PK2US - Pack Two Unsigned 16-bit Scalars
423 .. opcode:: PK4B - Pack Four Signed 8-bit Scalars
428 .. opcode:: PK4UB - Pack Four Unsigned 8-bit Scalars
433 .. opcode:: SEQ - Set On Equal
437 dst.x = (src0.x == src1.x) ? 1.0F : 0.0F
439 dst.y = (src0.y == src1.y) ? 1.0F : 0.0F
441 dst.z = (src0.z == src1.z) ? 1.0F : 0.0F
443 dst.w = (src0.w == src1.w) ? 1.0F : 0.0F
446 .. opcode:: SGT - Set On Greater Than
450 dst.x = (src0.x > src1.x) ? 1.0F : 0.0F
452 dst.y = (src0.y > src1.y) ? 1.0F : 0.0F
454 dst.z = (src0.z > src1.z) ? 1.0F : 0.0F
456 dst.w = (src0.w > src1.w) ? 1.0F : 0.0F
459 .. opcode:: SIN - Sine
461 This instruction replicates its result.
468 .. opcode:: SLE - Set On Less Equal Than
472 dst.x = (src0.x <= src1.x) ? 1.0F : 0.0F
474 dst.y = (src0.y <= src1.y) ? 1.0F : 0.0F
476 dst.z = (src0.z <= src1.z) ? 1.0F : 0.0F
478 dst.w = (src0.w <= src1.w) ? 1.0F : 0.0F
481 .. opcode:: SNE - Set On Not Equal
485 dst.x = (src0.x != src1.x) ? 1.0F : 0.0F
487 dst.y = (src0.y != src1.y) ? 1.0F : 0.0F
489 dst.z = (src0.z != src1.z) ? 1.0F : 0.0F
491 dst.w = (src0.w != src1.w) ? 1.0F : 0.0F
494 .. opcode:: TEX - Texture Lookup
496 for array textures src0.y contains the slice for 1D,
497 and src0.z contain the slice for 2D.
499 for shadow textures with no arrays (and not cube map),
500 src0.z contains the reference value.
502 for shadow textures with arrays, src0.z contains
503 the reference value for 1D arrays, and src0.w contains
504 the reference value for 2D arrays and cube maps.
506 for cube map array shadow textures, the reference value
507 cannot be passed in src0.w, and TEX2 must be used instead.
513 shadow_ref = src0.z or src0.w (optional)
517 dst = texture\_sample(unit, coord, shadow_ref)
520 .. opcode:: TEX2 - Texture Lookup (for shadow cube map arrays only)
522 this is the same as TEX, but uses another reg to encode the
533 dst = texture\_sample(unit, coord, shadow_ref)
538 .. opcode:: TXD - Texture Lookup with Derivatives
550 dst = texture\_sample\_deriv(unit, coord, ddx, ddy)
553 .. opcode:: TXP - Projective Texture Lookup
557 coord.x = src0.x / src0.w
559 coord.y = src0.y / src0.w
561 coord.z = src0.z / src0.w
567 dst = texture\_sample(unit, coord)
570 .. opcode:: UP2H - Unpack Two 16-Bit Floats
574 dst.x = f16\_to\_f32(src0.x \& 0xffff)
576 dst.y = f16\_to\_f32(src0.x >> 16)
578 dst.z = f16\_to\_f32(src0.x \& 0xffff)
580 dst.w = f16\_to\_f32(src0.x >> 16)
584 Considered for removal.
586 .. opcode:: UP2US - Unpack Two Unsigned 16-Bit Scalars
592 Considered for removal.
594 .. opcode:: UP4B - Unpack Four Signed 8-Bit Values
600 Considered for removal.
602 .. opcode:: UP4UB - Unpack Four Unsigned 8-Bit Scalars
608 Considered for removal.
611 .. opcode:: ARR - Address Register Load With Round
615 dst.x = (int) round(src.x)
617 dst.y = (int) round(src.y)
619 dst.z = (int) round(src.z)
621 dst.w = (int) round(src.w)
624 .. opcode:: SSG - Set Sign
628 dst.x = (src.x > 0) ? 1 : (src.x < 0) ? -1 : 0
630 dst.y = (src.y > 0) ? 1 : (src.y < 0) ? -1 : 0
632 dst.z = (src.z > 0) ? 1 : (src.z < 0) ? -1 : 0
634 dst.w = (src.w > 0) ? 1 : (src.w < 0) ? -1 : 0
637 .. opcode:: CMP - Compare
641 dst.x = (src0.x < 0) ? src1.x : src2.x
643 dst.y = (src0.y < 0) ? src1.y : src2.y
645 dst.z = (src0.z < 0) ? src1.z : src2.z
647 dst.w = (src0.w < 0) ? src1.w : src2.w
650 .. opcode:: KILL_IF - Conditional Discard
652 Conditional discard. Allowed in fragment shaders only.
656 if (src.x < 0 || src.y < 0 || src.z < 0 || src.w < 0)
661 .. opcode:: KILL - Discard
663 Unconditional discard. Allowed in fragment shaders only.
666 .. opcode:: SCS - Sine Cosine
679 .. opcode:: TXB - Texture Lookup With Bias
681 for cube map array textures and shadow cube maps, the bias value
682 cannot be passed in src0.w, and TXB2 must be used instead.
684 if the target is a shadow texture, the reference value is always
685 in src.z (this prevents shadow 3d and shadow 2d arrays from
686 using this instruction, but this is not needed).
702 dst = texture\_sample(unit, coord, bias)
705 .. opcode:: TXB2 - Texture Lookup With Bias (some cube maps only)
707 this is the same as TXB, but uses another reg to encode the
708 lod bias value for cube map arrays and shadow cube maps.
709 Presumably shadow 2d arrays and shadow 3d targets could use
710 this encoding too, but this is not legal.
712 shadow cube map arrays are neither possible nor required.
722 dst = texture\_sample(unit, coord, bias)
725 .. opcode:: DIV - Divide
729 dst.x = \frac{src0.x}{src1.x}
731 dst.y = \frac{src0.y}{src1.y}
733 dst.z = \frac{src0.z}{src1.z}
735 dst.w = \frac{src0.w}{src1.w}
738 .. opcode:: DP2 - 2-component Dot Product
740 This instruction replicates its result.
744 dst = src0.x \times src1.x + src0.y \times src1.y
747 .. opcode:: TEX_LZ - Texture Lookup With LOD = 0
749 This is the same as TXL with LOD = 0. Like every texture opcode, it obeys
750 pipe_sampler_view::u.tex.first_level and pipe_sampler_state::min_lod.
751 There is no way to override those two in shaders.
767 dst = texture\_sample(unit, coord, lod)
770 .. opcode:: TXL - Texture Lookup With explicit LOD
772 for cube map array textures, the explicit lod value
773 cannot be passed in src0.w, and TXL2 must be used instead.
775 if the target is a shadow texture, the reference value is always
776 in src.z (this prevents shadow 3d / 2d array / cube targets from
777 using this instruction, but this is not needed).
793 dst = texture\_sample(unit, coord, lod)
796 .. opcode:: TXL2 - Texture Lookup With explicit LOD (for cube map arrays only)
798 this is the same as TXL, but uses another reg to encode the
800 Presumably shadow 3d / 2d array / cube targets could use
801 this encoding too, but this is not legal.
803 shadow cube map arrays are neither possible nor required.
813 dst = texture\_sample(unit, coord, lod)
817 ^^^^^^^^^^^^^^^^^^^^^^^^
819 These opcodes are primarily provided for special-use computational shaders.
820 Support for these opcodes indicated by a special pipe capability bit (TBD).
822 XXX doesn't look like most of the opcodes really belong here.
824 .. opcode:: CEIL - Ceiling
828 dst.x = \lceil src.x\rceil
830 dst.y = \lceil src.y\rceil
832 dst.z = \lceil src.z\rceil
834 dst.w = \lceil src.w\rceil
837 .. opcode:: TRUNC - Truncate
850 .. opcode:: MOD - Modulus
854 dst.x = src0.x \bmod src1.x
856 dst.y = src0.y \bmod src1.y
858 dst.z = src0.z \bmod src1.z
860 dst.w = src0.w \bmod src1.w
863 .. opcode:: UARL - Integer Address Register Load
865 Moves the contents of the source register, assumed to be an integer, into the
866 destination register, which is assumed to be an address (ADDR) register.
869 .. opcode:: TXF - Texel Fetch
871 As per NV_gpu_shader4, extract a single texel from a specified texture
872 image or PIPE_BUFFER resource. The source sampler may not be a CUBE or
874 four-component signed integer vector used to identify the single texel
875 accessed. 3 components + level. If the texture is multisampled, then
876 the fourth component indicates the sample, not the mipmap level.
877 Just like texture instructions, an optional
878 offset vector is provided, which is subject to various driver restrictions
879 (regarding range, source of offsets). This instruction ignores the sampler
882 TXF(uint_vec coord, int_vec offset).
885 .. opcode:: TXQ - Texture Size Query
887 As per NV_gpu_program4, retrieve the dimensions of the texture depending on
888 the target. For 1D (width), 2D/RECT/CUBE (width, height), 3D (width, height,
889 depth), 1D array (width, layers), 2D array (width, height, layers).
890 Also return the number of accessible levels (last_level - first_level + 1)
893 For components which don't return a resource dimension, their value
900 dst.x = texture\_width(unit, lod)
902 dst.y = texture\_height(unit, lod)
904 dst.z = texture\_depth(unit, lod)
906 dst.w = texture\_levels(unit)
909 .. opcode:: TXQS - Texture Samples Query
911 This retrieves the number of samples in the texture, and stores it
912 into the x component as an unsigned integer. The other components are
913 undefined. If the texture is not multisampled, this function returns
914 (1, undef, undef, undef).
918 dst.x = texture\_samples(unit)
921 .. opcode:: TG4 - Texture Gather
923 As per ARB_texture_gather, gathers the four texels to be used in a bi-linear
924 filtering operation and packs them into a single register. Only works with
925 2D, 2D array, cubemaps, and cubemaps arrays. For 2D textures, only the
926 addressing modes of the sampler and the top level of any mip pyramid are
927 used. Set W to zero. It behaves like the TEX instruction, but a filtered
928 sample is not generated. The four samples that contribute to filtering are
929 placed into xyzw in clockwise order, starting with the (u,v) texture
930 coordinate delta at the following locations (-, +), (+, +), (+, -), (-, -),
931 where the magnitude of the deltas are half a texel.
933 PIPE_CAP_TEXTURE_SM5 enhances this instruction to support shadow per-sample
934 depth compares, single component selection, and a non-constant offset. It
935 doesn't allow support for the GL independent offset to get i0,j0. This would
936 require another CAP is hw can do it natively. For now we lower that before
945 dst = texture\_gather4 (unit, coord, component)
947 (with SM5 - cube array shadow)
955 dst = texture\_gather (uint, coord, compare)
957 .. opcode:: LODQ - level of detail query
959 Compute the LOD information that the texture pipe would use to access the
960 texture. The Y component contains the computed LOD lambda_prime. The X
961 component contains the LOD that will be accessed, based on min/max lod's
968 dst.xy = lodq(uint, coord);
970 .. opcode:: CLOCK - retrieve the current shader time
972 Invoking this instruction multiple times in the same shader should
973 cause monotonically increasing values to be returned. The values
974 are implicitly 64-bit, so if fewer than 64 bits of precision are
975 available, to provide expected wraparound semantics, the value
976 should be shifted up so that the most significant bit of the time
977 is the most significant bit of the 64-bit value.
985 ^^^^^^^^^^^^^^^^^^^^^^^^
986 These opcodes are used for integer operations.
987 Support for these opcodes indicated by PIPE_SHADER_CAP_INTEGERS (all of them?)
990 .. opcode:: I2F - Signed Integer To Float
992 Rounding is unspecified (round to nearest even suggested).
996 dst.x = (float) src.x
998 dst.y = (float) src.y
1000 dst.z = (float) src.z
1002 dst.w = (float) src.w
1005 .. opcode:: U2F - Unsigned Integer To Float
1007 Rounding is unspecified (round to nearest even suggested).
1011 dst.x = (float) src.x
1013 dst.y = (float) src.y
1015 dst.z = (float) src.z
1017 dst.w = (float) src.w
1020 .. opcode:: F2I - Float to Signed Integer
1022 Rounding is towards zero (truncate).
1023 Values outside signed range (including NaNs) produce undefined results.
1036 .. opcode:: F2U - Float to Unsigned Integer
1038 Rounding is towards zero (truncate).
1039 Values outside unsigned range (including NaNs) produce undefined results.
1043 dst.x = (unsigned) src.x
1045 dst.y = (unsigned) src.y
1047 dst.z = (unsigned) src.z
1049 dst.w = (unsigned) src.w
1052 .. opcode:: UADD - Integer Add
1054 This instruction works the same for signed and unsigned integers.
1055 The low 32bit of the result is returned.
1059 dst.x = src0.x + src1.x
1061 dst.y = src0.y + src1.y
1063 dst.z = src0.z + src1.z
1065 dst.w = src0.w + src1.w
1068 .. opcode:: UMAD - Integer Multiply And Add
1070 This instruction works the same for signed and unsigned integers.
1071 The multiplication returns the low 32bit (as does the result itself).
1075 dst.x = src0.x \times src1.x + src2.x
1077 dst.y = src0.y \times src1.y + src2.y
1079 dst.z = src0.z \times src1.z + src2.z
1081 dst.w = src0.w \times src1.w + src2.w
1084 .. opcode:: UMUL - Integer Multiply
1086 This instruction works the same for signed and unsigned integers.
1087 The low 32bit of the result is returned.
1091 dst.x = src0.x \times src1.x
1093 dst.y = src0.y \times src1.y
1095 dst.z = src0.z \times src1.z
1097 dst.w = src0.w \times src1.w
1100 .. opcode:: IMUL_HI - Signed Integer Multiply High Bits
1102 The high 32bits of the multiplication of 2 signed integers are returned.
1106 dst.x = (src0.x \times src1.x) >> 32
1108 dst.y = (src0.y \times src1.y) >> 32
1110 dst.z = (src0.z \times src1.z) >> 32
1112 dst.w = (src0.w \times src1.w) >> 32
1115 .. opcode:: UMUL_HI - Unsigned Integer Multiply High Bits
1117 The high 32bits of the multiplication of 2 unsigned integers are returned.
1121 dst.x = (src0.x \times src1.x) >> 32
1123 dst.y = (src0.y \times src1.y) >> 32
1125 dst.z = (src0.z \times src1.z) >> 32
1127 dst.w = (src0.w \times src1.w) >> 32
1130 .. opcode:: IDIV - Signed Integer Division
1132 TBD: behavior for division by zero.
1136 dst.x = \frac{src0.x}{src1.x}
1138 dst.y = \frac{src0.y}{src1.y}
1140 dst.z = \frac{src0.z}{src1.z}
1142 dst.w = \frac{src0.w}{src1.w}
1145 .. opcode:: UDIV - Unsigned Integer Division
1147 For division by zero, 0xffffffff is returned.
1151 dst.x = \frac{src0.x}{src1.x}
1153 dst.y = \frac{src0.y}{src1.y}
1155 dst.z = \frac{src0.z}{src1.z}
1157 dst.w = \frac{src0.w}{src1.w}
1160 .. opcode:: UMOD - Unsigned Integer Remainder
1162 If second arg is zero, 0xffffffff is returned.
1166 dst.x = src0.x \bmod src1.x
1168 dst.y = src0.y \bmod src1.y
1170 dst.z = src0.z \bmod src1.z
1172 dst.w = src0.w \bmod src1.w
1175 .. opcode:: NOT - Bitwise Not
1188 .. opcode:: AND - Bitwise And
1192 dst.x = src0.x \& src1.x
1194 dst.y = src0.y \& src1.y
1196 dst.z = src0.z \& src1.z
1198 dst.w = src0.w \& src1.w
1201 .. opcode:: OR - Bitwise Or
1205 dst.x = src0.x | src1.x
1207 dst.y = src0.y | src1.y
1209 dst.z = src0.z | src1.z
1211 dst.w = src0.w | src1.w
1214 .. opcode:: XOR - Bitwise Xor
1218 dst.x = src0.x \oplus src1.x
1220 dst.y = src0.y \oplus src1.y
1222 dst.z = src0.z \oplus src1.z
1224 dst.w = src0.w \oplus src1.w
1227 .. opcode:: IMAX - Maximum of Signed Integers
1231 dst.x = max(src0.x, src1.x)
1233 dst.y = max(src0.y, src1.y)
1235 dst.z = max(src0.z, src1.z)
1237 dst.w = max(src0.w, src1.w)
1240 .. opcode:: UMAX - Maximum of Unsigned Integers
1244 dst.x = max(src0.x, src1.x)
1246 dst.y = max(src0.y, src1.y)
1248 dst.z = max(src0.z, src1.z)
1250 dst.w = max(src0.w, src1.w)
1253 .. opcode:: IMIN - Minimum of Signed Integers
1257 dst.x = min(src0.x, src1.x)
1259 dst.y = min(src0.y, src1.y)
1261 dst.z = min(src0.z, src1.z)
1263 dst.w = min(src0.w, src1.w)
1266 .. opcode:: UMIN - Minimum of Unsigned Integers
1270 dst.x = min(src0.x, src1.x)
1272 dst.y = min(src0.y, src1.y)
1274 dst.z = min(src0.z, src1.z)
1276 dst.w = min(src0.w, src1.w)
1279 .. opcode:: SHL - Shift Left
1281 The shift count is masked with 0x1f before the shift is applied.
1285 dst.x = src0.x << (0x1f \& src1.x)
1287 dst.y = src0.y << (0x1f \& src1.y)
1289 dst.z = src0.z << (0x1f \& src1.z)
1291 dst.w = src0.w << (0x1f \& src1.w)
1294 .. opcode:: ISHR - Arithmetic Shift Right (of Signed Integer)
1296 The shift count is masked with 0x1f before the shift is applied.
1300 dst.x = src0.x >> (0x1f \& src1.x)
1302 dst.y = src0.y >> (0x1f \& src1.y)
1304 dst.z = src0.z >> (0x1f \& src1.z)
1306 dst.w = src0.w >> (0x1f \& src1.w)
1309 .. opcode:: USHR - Logical Shift Right
1311 The shift count is masked with 0x1f before the shift is applied.
1315 dst.x = src0.x >> (unsigned) (0x1f \& src1.x)
1317 dst.y = src0.y >> (unsigned) (0x1f \& src1.y)
1319 dst.z = src0.z >> (unsigned) (0x1f \& src1.z)
1321 dst.w = src0.w >> (unsigned) (0x1f \& src1.w)
1324 .. opcode:: UCMP - Integer Conditional Move
1328 dst.x = src0.x ? src1.x : src2.x
1330 dst.y = src0.y ? src1.y : src2.y
1332 dst.z = src0.z ? src1.z : src2.z
1334 dst.w = src0.w ? src1.w : src2.w
1338 .. opcode:: ISSG - Integer Set Sign
1342 dst.x = (src0.x < 0) ? -1 : (src0.x > 0) ? 1 : 0
1344 dst.y = (src0.y < 0) ? -1 : (src0.y > 0) ? 1 : 0
1346 dst.z = (src0.z < 0) ? -1 : (src0.z > 0) ? 1 : 0
1348 dst.w = (src0.w < 0) ? -1 : (src0.w > 0) ? 1 : 0
1352 .. opcode:: FSLT - Float Set On Less Than (ordered)
1354 Same comparison as SLT but returns integer instead of 1.0/0.0 float
1358 dst.x = (src0.x < src1.x) ? \sim 0 : 0
1360 dst.y = (src0.y < src1.y) ? \sim 0 : 0
1362 dst.z = (src0.z < src1.z) ? \sim 0 : 0
1364 dst.w = (src0.w < src1.w) ? \sim 0 : 0
1367 .. opcode:: ISLT - Signed Integer Set On Less Than
1371 dst.x = (src0.x < src1.x) ? \sim 0 : 0
1373 dst.y = (src0.y < src1.y) ? \sim 0 : 0
1375 dst.z = (src0.z < src1.z) ? \sim 0 : 0
1377 dst.w = (src0.w < src1.w) ? \sim 0 : 0
1380 .. opcode:: USLT - Unsigned Integer Set On Less Than
1384 dst.x = (src0.x < src1.x) ? \sim 0 : 0
1386 dst.y = (src0.y < src1.y) ? \sim 0 : 0
1388 dst.z = (src0.z < src1.z) ? \sim 0 : 0
1390 dst.w = (src0.w < src1.w) ? \sim 0 : 0
1393 .. opcode:: FSGE - Float Set On Greater Equal Than (ordered)
1395 Same comparison as SGE but returns integer instead of 1.0/0.0 float
1399 dst.x = (src0.x >= src1.x) ? \sim 0 : 0
1401 dst.y = (src0.y >= src1.y) ? \sim 0 : 0
1403 dst.z = (src0.z >= src1.z) ? \sim 0 : 0
1405 dst.w = (src0.w >= src1.w) ? \sim 0 : 0
1408 .. opcode:: ISGE - Signed Integer Set On Greater Equal Than
1412 dst.x = (src0.x >= src1.x) ? \sim 0 : 0
1414 dst.y = (src0.y >= src1.y) ? \sim 0 : 0
1416 dst.z = (src0.z >= src1.z) ? \sim 0 : 0
1418 dst.w = (src0.w >= src1.w) ? \sim 0 : 0
1421 .. opcode:: USGE - Unsigned Integer Set On Greater Equal Than
1425 dst.x = (src0.x >= src1.x) ? \sim 0 : 0
1427 dst.y = (src0.y >= src1.y) ? \sim 0 : 0
1429 dst.z = (src0.z >= src1.z) ? \sim 0 : 0
1431 dst.w = (src0.w >= src1.w) ? \sim 0 : 0
1434 .. opcode:: FSEQ - Float Set On Equal (ordered)
1436 Same comparison as SEQ but returns integer instead of 1.0/0.0 float
1440 dst.x = (src0.x == src1.x) ? \sim 0 : 0
1442 dst.y = (src0.y == src1.y) ? \sim 0 : 0
1444 dst.z = (src0.z == src1.z) ? \sim 0 : 0
1446 dst.w = (src0.w == src1.w) ? \sim 0 : 0
1449 .. opcode:: USEQ - Integer Set On Equal
1453 dst.x = (src0.x == src1.x) ? \sim 0 : 0
1455 dst.y = (src0.y == src1.y) ? \sim 0 : 0
1457 dst.z = (src0.z == src1.z) ? \sim 0 : 0
1459 dst.w = (src0.w == src1.w) ? \sim 0 : 0
1462 .. opcode:: FSNE - Float Set On Not Equal (unordered)
1464 Same comparison as SNE but returns integer instead of 1.0/0.0 float
1468 dst.x = (src0.x != src1.x) ? \sim 0 : 0
1470 dst.y = (src0.y != src1.y) ? \sim 0 : 0
1472 dst.z = (src0.z != src1.z) ? \sim 0 : 0
1474 dst.w = (src0.w != src1.w) ? \sim 0 : 0
1477 .. opcode:: USNE - Integer Set On Not Equal
1481 dst.x = (src0.x != src1.x) ? \sim 0 : 0
1483 dst.y = (src0.y != src1.y) ? \sim 0 : 0
1485 dst.z = (src0.z != src1.z) ? \sim 0 : 0
1487 dst.w = (src0.w != src1.w) ? \sim 0 : 0
1490 .. opcode:: INEG - Integer Negate
1505 .. opcode:: IABS - Integer Absolute Value
1519 These opcodes are used for bit-level manipulation of integers.
1521 .. opcode:: IBFE - Signed Bitfield Extract
1523 Like GLSL bitfieldExtract. Extracts a set of bits from the input, and
1524 sign-extends them if the high bit of the extracted window is set.
1528 def ibfe(value, offset, bits):
1529 if offset < 0 or bits < 0 or offset + bits > 32:
1531 if bits == 0: return 0
1532 # Note: >> sign-extends
1533 return (value << (32 - offset - bits)) >> (32 - bits)
1535 .. opcode:: UBFE - Unsigned Bitfield Extract
1537 Like GLSL bitfieldExtract. Extracts a set of bits from the input, without
1542 def ubfe(value, offset, bits):
1543 if offset < 0 or bits < 0 or offset + bits > 32:
1545 if bits == 0: return 0
1546 # Note: >> does not sign-extend
1547 return (value << (32 - offset - bits)) >> (32 - bits)
1549 .. opcode:: BFI - Bitfield Insert
1551 Like GLSL bitfieldInsert. Replaces a bit region of 'base' with the low bits
1556 def bfi(base, insert, offset, bits):
1557 if offset < 0 or bits < 0 or offset + bits > 32:
1559 # << defined such that mask == ~0 when bits == 32, offset == 0
1560 mask = ((1 << bits) - 1) << offset
1561 return ((insert << offset) & mask) | (base & ~mask)
1563 .. opcode:: BREV - Bitfield Reverse
1565 See SM5 instruction BFREV. Reverses the bits of the argument.
1567 .. opcode:: POPC - Population Count
1569 See SM5 instruction COUNTBITS. Counts the number of set bits in the argument.
1571 .. opcode:: LSB - Index of lowest set bit
1573 See SM5 instruction FIRSTBIT_LO. Computes the 0-based index of the first set
1574 bit of the argument. Returns -1 if none are set.
1576 .. opcode:: IMSB - Index of highest non-sign bit
1578 See SM5 instruction FIRSTBIT_SHI. Computes the 0-based index of the highest
1579 non-sign bit of the argument (i.e. highest 0 bit for negative numbers,
1580 highest 1 bit for positive numbers). Returns -1 if all bits are the same
1581 (i.e. for inputs 0 and -1).
1583 .. opcode:: UMSB - Index of highest set bit
1585 See SM5 instruction FIRSTBIT_HI. Computes the 0-based index of the highest
1586 set bit of the argument. Returns -1 if none are set.
1589 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1591 These opcodes are only supported in geometry shaders; they have no meaning
1592 in any other type of shader.
1594 .. opcode:: EMIT - Emit
1596 Generate a new vertex for the current primitive into the specified vertex
1597 stream using the values in the output registers.
1600 .. opcode:: ENDPRIM - End Primitive
1602 Complete the current primitive in the specified vertex stream (consisting of
1603 the emitted vertices), and start a new one.
1609 These opcodes are part of :term:`GLSL`'s opcode set. Support for these
1610 opcodes is determined by a special capability bit, ``GLSL``.
1611 Some require glsl version 1.30 (UIF/BREAKC/SWITCH/CASE/DEFAULT/ENDSWITCH).
1613 .. opcode:: CAL - Subroutine Call
1619 .. opcode:: RET - Subroutine Call Return
1624 .. opcode:: CONT - Continue
1626 Unconditionally moves the point of execution to the instruction after the
1627 last bgnloop. The instruction must appear within a bgnloop/endloop.
1631 Support for CONT is determined by a special capability bit,
1632 ``TGSI_CONT_SUPPORTED``. See :ref:`Screen` for more information.
1635 .. opcode:: BGNLOOP - Begin a Loop
1637 Start a loop. Must have a matching endloop.
1640 .. opcode:: BGNSUB - Begin Subroutine
1642 Starts definition of a subroutine. Must have a matching endsub.
1645 .. opcode:: ENDLOOP - End a Loop
1647 End a loop started with bgnloop.
1650 .. opcode:: ENDSUB - End Subroutine
1652 Ends definition of a subroutine.
1655 .. opcode:: NOP - No Operation
1660 .. opcode:: BRK - Break
1662 Unconditionally moves the point of execution to the instruction after the
1663 next endloop or endswitch. The instruction must appear within a loop/endloop
1664 or switch/endswitch.
1667 .. opcode:: BREAKC - Break Conditional
1669 Conditionally moves the point of execution to the instruction after the
1670 next endloop or endswitch. The instruction must appear within a loop/endloop
1671 or switch/endswitch.
1672 Condition evaluates to true if src0.x != 0 where src0.x is interpreted
1673 as an integer register.
1677 Considered for removal as it's quite inconsistent wrt other opcodes
1678 (could emulate with UIF/BRK/ENDIF).
1681 .. opcode:: IF - Float If
1683 Start an IF ... ELSE .. ENDIF block. Condition evaluates to true if
1687 where src0.x is interpreted as a floating point register.
1690 .. opcode:: UIF - Bitwise If
1692 Start an UIF ... ELSE .. ENDIF block. Condition evaluates to true if
1696 where src0.x is interpreted as an integer register.
1699 .. opcode:: ELSE - Else
1701 Starts an else block, after an IF or UIF statement.
1704 .. opcode:: ENDIF - End If
1706 Ends an IF or UIF block.
1709 .. opcode:: SWITCH - Switch
1711 Starts a C-style switch expression. The switch consists of one or multiple
1712 CASE statements, and at most one DEFAULT statement. Execution of a statement
1713 ends when a BRK is hit, but just like in C falling through to other cases
1714 without a break is allowed. Similarly, DEFAULT label is allowed anywhere not
1715 just as last statement, and fallthrough is allowed into/from it.
1716 CASE src arguments are evaluated at bit level against the SWITCH src argument.
1722 (some instructions here)
1725 (some instructions here)
1728 (some instructions here)
1733 .. opcode:: CASE - Switch case
1735 This represents a switch case label. The src arg must be an integer immediate.
1738 .. opcode:: DEFAULT - Switch default
1740 This represents the default case in the switch, which is taken if no other
1744 .. opcode:: ENDSWITCH - End of switch
1746 Ends a switch expression.
1752 The interpolation instructions allow an input to be interpolated in a
1753 different way than its declaration. This corresponds to the GLSL 4.00
1754 interpolateAt* functions. The first argument of each of these must come from
1755 ``TGSI_FILE_INPUT``.
1757 .. opcode:: INTERP_CENTROID - Interpolate at the centroid
1759 Interpolates the varying specified by src0 at the centroid
1761 .. opcode:: INTERP_SAMPLE - Interpolate at the specified sample
1763 Interpolates the varying specified by src0 at the sample id specified by
1764 src1.x (interpreted as an integer)
1766 .. opcode:: INTERP_OFFSET - Interpolate at the specified offset
1768 Interpolates the varying specified by src0 at the offset src1.xy from the
1769 pixel center (interpreted as floats)
1777 The double-precision opcodes reinterpret four-component vectors into
1778 two-component vectors with doubled precision in each component.
1780 .. opcode:: DABS - Absolute
1788 .. opcode:: DADD - Add
1792 dst.xy = src0.xy + src1.xy
1794 dst.zw = src0.zw + src1.zw
1796 .. opcode:: DSEQ - Set on Equal
1800 dst.x = src0.xy == src1.xy ? \sim 0 : 0
1802 dst.z = src0.zw == src1.zw ? \sim 0 : 0
1804 .. opcode:: DSNE - Set on Equal
1808 dst.x = src0.xy != src1.xy ? \sim 0 : 0
1810 dst.z = src0.zw != src1.zw ? \sim 0 : 0
1812 .. opcode:: DSLT - Set on Less than
1816 dst.x = src0.xy < src1.xy ? \sim 0 : 0
1818 dst.z = src0.zw < src1.zw ? \sim 0 : 0
1820 .. opcode:: DSGE - Set on Greater equal
1824 dst.x = src0.xy >= src1.xy ? \sim 0 : 0
1826 dst.z = src0.zw >= src1.zw ? \sim 0 : 0
1828 .. opcode:: DFRAC - Fraction
1832 dst.xy = src.xy - \lfloor src.xy\rfloor
1834 dst.zw = src.zw - \lfloor src.zw\rfloor
1836 .. opcode:: DTRUNC - Truncate
1840 dst.xy = trunc(src.xy)
1842 dst.zw = trunc(src.zw)
1844 .. opcode:: DCEIL - Ceiling
1848 dst.xy = \lceil src.xy\rceil
1850 dst.zw = \lceil src.zw\rceil
1852 .. opcode:: DFLR - Floor
1856 dst.xy = \lfloor src.xy\rfloor
1858 dst.zw = \lfloor src.zw\rfloor
1860 .. opcode:: DROUND - Fraction
1864 dst.xy = round(src.xy)
1866 dst.zw = round(src.zw)
1868 .. opcode:: DSSG - Set Sign
1872 dst.xy = (src.xy > 0) ? 1.0 : (src.xy < 0) ? -1.0 : 0.0
1874 dst.zw = (src.zw > 0) ? 1.0 : (src.zw < 0) ? -1.0 : 0.0
1876 .. opcode:: DFRACEXP - Convert Number to Fractional and Integral Components
1878 Like the ``frexp()`` routine in many math libraries, this opcode stores the
1879 exponent of its source to ``dst0``, and the significand to ``dst1``, such that
1880 :math:`dst1 \times 2^{dst0} = src` .
1884 dst0.xy = exp(src.xy)
1886 dst1.xy = frac(src.xy)
1888 dst0.zw = exp(src.zw)
1890 dst1.zw = frac(src.zw)
1892 .. opcode:: DLDEXP - Multiply Number by Integral Power of 2
1894 This opcode is the inverse of :opcode:`DFRACEXP`. The second
1895 source is an integer.
1899 dst.xy = src0.xy \times 2^{src1.x}
1901 dst.zw = src0.zw \times 2^{src1.y}
1903 .. opcode:: DMIN - Minimum
1907 dst.xy = min(src0.xy, src1.xy)
1909 dst.zw = min(src0.zw, src1.zw)
1911 .. opcode:: DMAX - Maximum
1915 dst.xy = max(src0.xy, src1.xy)
1917 dst.zw = max(src0.zw, src1.zw)
1919 .. opcode:: DMUL - Multiply
1923 dst.xy = src0.xy \times src1.xy
1925 dst.zw = src0.zw \times src1.zw
1928 .. opcode:: DMAD - Multiply And Add
1932 dst.xy = src0.xy \times src1.xy + src2.xy
1934 dst.zw = src0.zw \times src1.zw + src2.zw
1937 .. opcode:: DFMA - Fused Multiply-Add
1939 Perform a * b + c with no intermediate rounding step.
1943 dst.xy = src0.xy \times src1.xy + src2.xy
1945 dst.zw = src0.zw \times src1.zw + src2.zw
1948 .. opcode:: DDIV - Divide
1952 dst.xy = \frac{src0.xy}{src1.xy}
1954 dst.zw = \frac{src0.zw}{src1.zw}
1957 .. opcode:: DRCP - Reciprocal
1961 dst.xy = \frac{1}{src.xy}
1963 dst.zw = \frac{1}{src.zw}
1965 .. opcode:: DSQRT - Square Root
1969 dst.xy = \sqrt{src.xy}
1971 dst.zw = \sqrt{src.zw}
1973 .. opcode:: DRSQ - Reciprocal Square Root
1977 dst.xy = \frac{1}{\sqrt{src.xy}}
1979 dst.zw = \frac{1}{\sqrt{src.zw}}
1981 .. opcode:: F2D - Float to Double
1985 dst.xy = double(src0.x)
1987 dst.zw = double(src0.y)
1989 .. opcode:: D2F - Double to Float
1993 dst.x = float(src0.xy)
1995 dst.y = float(src0.zw)
1997 .. opcode:: I2D - Int to Double
2001 dst.xy = double(src0.x)
2003 dst.zw = double(src0.y)
2005 .. opcode:: D2I - Double to Int
2009 dst.x = int(src0.xy)
2011 dst.y = int(src0.zw)
2013 .. opcode:: U2D - Unsigned Int to Double
2017 dst.xy = double(src0.x)
2019 dst.zw = double(src0.y)
2021 .. opcode:: D2U - Double to Unsigned Int
2025 dst.x = unsigned(src0.xy)
2027 dst.y = unsigned(src0.zw)
2032 The 64-bit integer opcodes reinterpret four-component vectors into
2033 two-component vectors with 64-bits in each component.
2035 .. opcode:: I64ABS - 64-bit Integer Absolute Value
2043 .. opcode:: I64NEG - 64-bit Integer Negate
2053 .. opcode:: I64SSG - 64-bit Integer Set Sign
2057 dst.xy = (src0.xy < 0) ? -1 : (src0.xy > 0) ? 1 : 0
2059 dst.zw = (src0.zw < 0) ? -1 : (src0.zw > 0) ? 1 : 0
2061 .. opcode:: U64ADD - 64-bit Integer Add
2065 dst.xy = src0.xy + src1.xy
2067 dst.zw = src0.zw + src1.zw
2069 .. opcode:: U64MUL - 64-bit Integer Multiply
2073 dst.xy = src0.xy * src1.xy
2075 dst.zw = src0.zw * src1.zw
2077 .. opcode:: U64SEQ - 64-bit Integer Set on Equal
2081 dst.x = src0.xy == src1.xy ? \sim 0 : 0
2083 dst.z = src0.zw == src1.zw ? \sim 0 : 0
2085 .. opcode:: U64SNE - 64-bit Integer Set on Not Equal
2089 dst.x = src0.xy != src1.xy ? \sim 0 : 0
2091 dst.z = src0.zw != src1.zw ? \sim 0 : 0
2093 .. opcode:: U64SLT - 64-bit Unsigned Integer Set on Less Than
2097 dst.x = src0.xy < src1.xy ? \sim 0 : 0
2099 dst.z = src0.zw < src1.zw ? \sim 0 : 0
2101 .. opcode:: U64SGE - 64-bit Unsigned Integer Set on Greater Equal
2105 dst.x = src0.xy >= src1.xy ? \sim 0 : 0
2107 dst.z = src0.zw >= src1.zw ? \sim 0 : 0
2109 .. opcode:: I64SLT - 64-bit Signed Integer Set on Less Than
2113 dst.x = src0.xy < src1.xy ? \sim 0 : 0
2115 dst.z = src0.zw < src1.zw ? \sim 0 : 0
2117 .. opcode:: I64SGE - 64-bit Signed Integer Set on Greater Equal
2121 dst.x = src0.xy >= src1.xy ? \sim 0 : 0
2123 dst.z = src0.zw >= src1.zw ? \sim 0 : 0
2125 .. opcode:: I64MIN - Minimum of 64-bit Signed Integers
2129 dst.xy = min(src0.xy, src1.xy)
2131 dst.zw = min(src0.zw, src1.zw)
2133 .. opcode:: U64MIN - Minimum of 64-bit Unsigned Integers
2137 dst.xy = min(src0.xy, src1.xy)
2139 dst.zw = min(src0.zw, src1.zw)
2141 .. opcode:: I64MAX - Maximum of 64-bit Signed Integers
2145 dst.xy = max(src0.xy, src1.xy)
2147 dst.zw = max(src0.zw, src1.zw)
2149 .. opcode:: U64MAX - Maximum of 64-bit Unsigned Integers
2153 dst.xy = max(src0.xy, src1.xy)
2155 dst.zw = max(src0.zw, src1.zw)
2157 .. opcode:: U64SHL - Shift Left 64-bit Unsigned Integer
2159 The shift count is masked with 0x3f before the shift is applied.
2163 dst.xy = src0.xy << (0x3f \& src1.x)
2165 dst.zw = src0.zw << (0x3f \& src1.y)
2167 .. opcode:: I64SHR - Arithmetic Shift Right (of 64-bit Signed Integer)
2169 The shift count is masked with 0x3f before the shift is applied.
2173 dst.xy = src0.xy >> (0x3f \& src1.x)
2175 dst.zw = src0.zw >> (0x3f \& src1.y)
2177 .. opcode:: U64SHR - Logical Shift Right (of 64-bit Unsigned Integer)
2179 The shift count is masked with 0x3f before the shift is applied.
2183 dst.xy = src0.xy >> (unsigned) (0x3f \& src1.x)
2185 dst.zw = src0.zw >> (unsigned) (0x3f \& src1.y)
2187 .. opcode:: I64DIV - 64-bit Signed Integer Division
2191 dst.xy = \frac{src0.xy}{src1.xy}
2193 dst.zw = \frac{src0.zw}{src1.zw}
2195 .. opcode:: U64DIV - 64-bit Unsigned Integer Division
2199 dst.xy = \frac{src0.xy}{src1.xy}
2201 dst.zw = \frac{src0.zw}{src1.zw}
2203 .. opcode:: U64MOD - 64-bit Unsigned Integer Remainder
2207 dst.xy = src0.xy \bmod src1.xy
2209 dst.zw = src0.zw \bmod src1.zw
2211 .. opcode:: I64MOD - 64-bit Signed Integer Remainder
2215 dst.xy = src0.xy \bmod src1.xy
2217 dst.zw = src0.zw \bmod src1.zw
2219 .. opcode:: F2U64 - Float to 64-bit Unsigned Int
2223 dst.xy = (uint64_t) src0.x
2225 dst.zw = (uint64_t) src0.y
2227 .. opcode:: F2I64 - Float to 64-bit Int
2231 dst.xy = (int64_t) src0.x
2233 dst.zw = (int64_t) src0.y
2235 .. opcode:: U2I64 - Unsigned Integer to 64-bit Integer
2237 This is a zero extension.
2241 dst.xy = (uint64_t) src0.x
2243 dst.zw = (uint64_t) src0.y
2245 .. opcode:: I2I64 - Signed Integer to 64-bit Integer
2247 This is a sign extension.
2251 dst.xy = (int64_t) src0.x
2253 dst.zw = (int64_t) src0.y
2255 .. opcode:: D2U64 - Double to 64-bit Unsigned Int
2259 dst.xy = (uint64_t) src0.xy
2261 dst.zw = (uint64_t) src0.zw
2263 .. opcode:: D2I64 - Double to 64-bit Int
2267 dst.xy = (int64_t) src0.xy
2269 dst.zw = (int64_t) src0.zw
2271 .. opcode:: U642F - 64-bit unsigned integer to float
2275 dst.x = (float) src0.xy
2277 dst.y = (float) src0.zw
2279 .. opcode:: I642F - 64-bit Int to Float
2283 dst.x = (float) src0.xy
2285 dst.y = (float) src0.zw
2287 .. opcode:: U642D - 64-bit unsigned integer to double
2291 dst.xy = (double) src0.xy
2293 dst.zw = (double) src0.zw
2295 .. opcode:: I642D - 64-bit Int to double
2299 dst.xy = (double) src0.xy
2301 dst.zw = (double) src0.zw
2303 .. _samplingopcodes:
2305 Resource Sampling Opcodes
2306 ^^^^^^^^^^^^^^^^^^^^^^^^^
2308 Those opcodes follow very closely semantics of the respective Direct3D
2309 instructions. If in doubt double check Direct3D documentation.
2310 Note that the swizzle on SVIEW (src1) determines texel swizzling
2315 Using provided address, sample data from the specified texture using the
2316 filtering mode identified by the given sampler. The source data may come from
2317 any resource type other than buffers.
2319 Syntax: ``SAMPLE dst, address, sampler_view, sampler``
2321 Example: ``SAMPLE TEMP[0], TEMP[1], SVIEW[0], SAMP[0]``
2323 .. opcode:: SAMPLE_I
2325 Simplified alternative to the SAMPLE instruction. Using the provided
2326 integer address, SAMPLE_I fetches data from the specified sampler view
2327 without any filtering. The source data may come from any resource type
2330 Syntax: ``SAMPLE_I dst, address, sampler_view``
2332 Example: ``SAMPLE_I TEMP[0], TEMP[1], SVIEW[0]``
2334 The 'address' is specified as unsigned integers. If the 'address' is out of
2335 range [0...(# texels - 1)] the result of the fetch is always 0 in all
2336 components. As such the instruction doesn't honor address wrap modes, in
2337 cases where that behavior is desirable 'SAMPLE' instruction should be used.
2338 address.w always provides an unsigned integer mipmap level. If the value is
2339 out of the range then the instruction always returns 0 in all components.
2340 address.yz are ignored for buffers and 1d textures. address.z is ignored
2341 for 1d texture arrays and 2d textures.
2343 For 1D texture arrays address.y provides the array index (also as unsigned
2344 integer). If the value is out of the range of available array indices
2345 [0... (array size - 1)] then the opcode always returns 0 in all components.
2346 For 2D texture arrays address.z provides the array index, otherwise it
2347 exhibits the same behavior as in the case for 1D texture arrays. The exact
2348 semantics of the source address are presented in the table below:
2350 +---------------------------+----+-----+-----+---------+
2351 | resource type | X | Y | Z | W |
2352 +===========================+====+=====+=====+=========+
2353 | ``PIPE_BUFFER`` | x | | | ignored |
2354 +---------------------------+----+-----+-----+---------+
2355 | ``PIPE_TEXTURE_1D`` | x | | | mpl |
2356 +---------------------------+----+-----+-----+---------+
2357 | ``PIPE_TEXTURE_2D`` | x | y | | mpl |
2358 +---------------------------+----+-----+-----+---------+
2359 | ``PIPE_TEXTURE_3D`` | x | y | z | mpl |
2360 +---------------------------+----+-----+-----+---------+
2361 | ``PIPE_TEXTURE_RECT`` | x | y | | mpl |
2362 +---------------------------+----+-----+-----+---------+
2363 | ``PIPE_TEXTURE_CUBE`` | not allowed as source |
2364 +---------------------------+----+-----+-----+---------+
2365 | ``PIPE_TEXTURE_1D_ARRAY`` | x | idx | | mpl |
2366 +---------------------------+----+-----+-----+---------+
2367 | ``PIPE_TEXTURE_2D_ARRAY`` | x | y | idx | mpl |
2368 +---------------------------+----+-----+-----+---------+
2370 Where 'mpl' is a mipmap level and 'idx' is the array index.
2372 .. opcode:: SAMPLE_I_MS
2374 Just like SAMPLE_I but allows fetch data from multi-sampled surfaces.
2376 Syntax: ``SAMPLE_I_MS dst, address, sampler_view, sample``
2378 .. opcode:: SAMPLE_B
2380 Just like the SAMPLE instruction with the exception that an additional bias
2381 is applied to the level of detail computed as part of the instruction
2384 Syntax: ``SAMPLE_B dst, address, sampler_view, sampler, lod_bias``
2386 Example: ``SAMPLE_B TEMP[0], TEMP[1], SVIEW[0], SAMP[0], TEMP[2].x``
2388 .. opcode:: SAMPLE_C
2390 Similar to the SAMPLE instruction but it performs a comparison filter. The
2391 operands to SAMPLE_C are identical to SAMPLE, except that there is an
2392 additional float32 operand, reference value, which must be a register with
2393 single-component, or a scalar literal. SAMPLE_C makes the hardware use the
2394 current samplers compare_func (in pipe_sampler_state) to compare reference
2395 value against the red component value for the surce resource at each texel
2396 that the currently configured texture filter covers based on the provided
2399 Syntax: ``SAMPLE_C dst, address, sampler_view.r, sampler, ref_value``
2401 Example: ``SAMPLE_C TEMP[0], TEMP[1], SVIEW[0].r, SAMP[0], TEMP[2].x``
2403 .. opcode:: SAMPLE_C_LZ
2405 Same as SAMPLE_C, but LOD is 0 and derivatives are ignored. The LZ stands
2408 Syntax: ``SAMPLE_C_LZ dst, address, sampler_view.r, sampler, ref_value``
2410 Example: ``SAMPLE_C_LZ TEMP[0], TEMP[1], SVIEW[0].r, SAMP[0], TEMP[2].x``
2413 .. opcode:: SAMPLE_D
2415 SAMPLE_D is identical to the SAMPLE opcode except that the derivatives for
2416 the source address in the x direction and the y direction are provided by
2419 Syntax: ``SAMPLE_D dst, address, sampler_view, sampler, der_x, der_y``
2421 Example: ``SAMPLE_D TEMP[0], TEMP[1], SVIEW[0], SAMP[0], TEMP[2], TEMP[3]``
2423 .. opcode:: SAMPLE_L
2425 SAMPLE_L is identical to the SAMPLE opcode except that the LOD is provided
2426 directly as a scalar value, representing no anisotropy.
2428 Syntax: ``SAMPLE_L dst, address, sampler_view, sampler, explicit_lod``
2430 Example: ``SAMPLE_L TEMP[0], TEMP[1], SVIEW[0], SAMP[0], TEMP[2].x``
2434 Gathers the four texels to be used in a bi-linear filtering operation and
2435 packs them into a single register. Only works with 2D, 2D array, cubemaps,
2436 and cubemaps arrays. For 2D textures, only the addressing modes of the
2437 sampler and the top level of any mip pyramid are used. Set W to zero. It
2438 behaves like the SAMPLE instruction, but a filtered sample is not
2439 generated. The four samples that contribute to filtering are placed into
2440 xyzw in counter-clockwise order, starting with the (u,v) texture coordinate
2441 delta at the following locations (-, +), (+, +), (+, -), (-, -), where the
2442 magnitude of the deltas are half a texel.
2445 .. opcode:: SVIEWINFO
2447 Query the dimensions of a given sampler view. dst receives width, height,
2448 depth or array size and number of mipmap levels as int4. The dst can have a
2449 writemask which will specify what info is the caller interested in.
2451 Syntax: ``SVIEWINFO dst, src_mip_level, sampler_view``
2453 Example: ``SVIEWINFO TEMP[0], TEMP[1].x, SVIEW[0]``
2455 src_mip_level is an unsigned integer scalar. If it's out of range then
2456 returns 0 for width, height and depth/array size but the total number of
2457 mipmap is still returned correctly for the given sampler view. The returned
2458 width, height and depth values are for the mipmap level selected by the
2459 src_mip_level and are in the number of texels. For 1d texture array width
2460 is in dst.x, array size is in dst.y and dst.z is 0. The number of mipmaps is
2461 still in dst.w. In contrast to d3d10 resinfo, there's no way in the tgsi
2462 instruction encoding to specify the return type (float/rcpfloat/uint), hence
2463 always using uint. Also, unlike the SAMPLE instructions, the swizzle on src1
2464 resinfo allowing swizzling dst values is ignored (due to the interaction
2465 with rcpfloat modifier which requires some swizzle handling in the state
2468 .. opcode:: SAMPLE_POS
2470 Query the position of a sample in the given resource or render target
2471 when per-sample fragment shading is in effect.
2473 Syntax: ``SAMPLE_POS dst, source, sample_index``
2475 dst receives float4 (x, y, undef, undef) indicated where the sample is
2476 located. Sample locations are in the range [0, 1] where 0.5 is the center
2479 source is either a sampler view (to indicate a shader resource) or temp
2480 register (to indicate the render target). The source register may have
2481 an optional swizzle to apply to the returned result
2483 sample_index is an integer scalar indicating which sample position is to
2486 If per-sample shading is not in effect or the source resource or render
2487 target is not multisampled, the result is (0.5, 0.5, undef, undef).
2489 NOTE: no driver has implemented this opcode yet (and no state tracker
2490 emits it). This information is subject to change.
2492 .. opcode:: SAMPLE_INFO
2494 Query the number of samples in a multisampled resource or render target.
2496 Syntax: ``SAMPLE_INFO dst, source``
2498 dst receives int4 (n, 0, 0, 0) where n is the number of samples in a
2499 resource or the render target.
2501 source is either a sampler view (to indicate a shader resource) or temp
2502 register (to indicate the render target). The source register may have
2503 an optional swizzle to apply to the returned result
2505 If per-sample shading is not in effect or the source resource or render
2506 target is not multisampled, the result is (1, 0, 0, 0).
2508 NOTE: no driver has implemented this opcode yet (and no state tracker
2509 emits it). This information is subject to change.
2511 .. _resourceopcodes:
2513 Resource Access Opcodes
2514 ^^^^^^^^^^^^^^^^^^^^^^^
2516 For these opcodes, the resource can be a BUFFER, IMAGE, or MEMORY.
2518 .. opcode:: LOAD - Fetch data from a shader buffer or image
2520 Syntax: ``LOAD dst, resource, address``
2522 Example: ``LOAD TEMP[0], BUFFER[0], TEMP[1]``
2524 Using the provided integer address, LOAD fetches data
2525 from the specified buffer or texture without any
2528 The 'address' is specified as a vector of unsigned
2529 integers. If the 'address' is out of range the result
2532 Only the first mipmap level of a resource can be read
2533 from using this instruction.
2535 For 1D or 2D texture arrays, the array index is
2536 provided as an unsigned integer in address.y or
2537 address.z, respectively. address.yz are ignored for
2538 buffers and 1D textures. address.z is ignored for 1D
2539 texture arrays and 2D textures. address.w is always
2542 A swizzle suffix may be added to the resource argument
2543 this will cause the resource data to be swizzled accordingly.
2545 .. opcode:: STORE - Write data to a shader resource
2547 Syntax: ``STORE resource, address, src``
2549 Example: ``STORE BUFFER[0], TEMP[0], TEMP[1]``
2551 Using the provided integer address, STORE writes data
2552 to the specified buffer or texture.
2554 The 'address' is specified as a vector of unsigned
2555 integers. If the 'address' is out of range the result
2558 Only the first mipmap level of a resource can be
2559 written to using this instruction.
2561 For 1D or 2D texture arrays, the array index is
2562 provided as an unsigned integer in address.y or
2563 address.z, respectively. address.yz are ignored for
2564 buffers and 1D textures. address.z is ignored for 1D
2565 texture arrays and 2D textures. address.w is always
2568 .. opcode:: RESQ - Query information about a resource
2570 Syntax: ``RESQ dst, resource``
2572 Example: ``RESQ TEMP[0], BUFFER[0]``
2574 Returns information about the buffer or image resource. For buffer
2575 resources, the size (in bytes) is returned in the x component. For
2576 image resources, .xyz will contain the width/height/layers of the
2577 image, while .w will contain the number of samples for multi-sampled
2580 .. opcode:: FBFETCH - Load data from framebuffer
2582 Syntax: ``FBFETCH dst, output``
2584 Example: ``FBFETCH TEMP[0], OUT[0]``
2586 This is only valid on ``COLOR`` semantic outputs. Returns the color
2587 of the current position in the framebuffer from before this fragment
2588 shader invocation. May return the same value from multiple calls for
2589 a particular output within a single invocation. Note that result may
2590 be undefined if a fragment is drawn multiple times without a blend
2594 .. _threadsyncopcodes:
2596 Inter-thread synchronization opcodes
2597 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
2599 These opcodes are intended for communication between threads running
2600 within the same compute grid. For now they're only valid in compute
2603 .. opcode:: BARRIER - Thread group barrier
2607 This opcode suspends the execution of the current thread until all
2608 the remaining threads in the working group reach the same point of
2609 the program. Results are unspecified if any of the remaining
2610 threads terminates or never reaches an executed BARRIER instruction.
2612 .. opcode:: MEMBAR - Memory barrier
2616 This opcode waits for the completion of all memory accesses based on
2617 the type passed in. The type is an immediate bitfield with the following
2620 Bit 0: Shader storage buffers
2621 Bit 1: Atomic buffers
2623 Bit 3: Shared memory
2626 These may be passed in in any combination. An implementation is free to not
2627 distinguish between these as it sees fit. However these map to all the
2628 possibilities made available by GLSL.
2635 These opcodes provide atomic variants of some common arithmetic and
2636 logical operations. In this context atomicity means that another
2637 concurrent memory access operation that affects the same memory
2638 location is guaranteed to be performed strictly before or after the
2639 entire execution of the atomic operation. The resource may be a BUFFER,
2640 IMAGE, or MEMORY. In the case of an image, the offset works the same as for
2641 ``LOAD`` and ``STORE``, specified above. These atomic operations may
2642 only be used with 32-bit integer image formats.
2644 .. opcode:: ATOMUADD - Atomic integer addition
2646 Syntax: ``ATOMUADD dst, resource, offset, src``
2648 Example: ``ATOMUADD TEMP[0], BUFFER[0], TEMP[1], TEMP[2]``
2650 The following operation is performed atomically:
2654 dst_x = resource[offset]
2656 resource[offset] = dst_x + src_x
2659 .. opcode:: ATOMXCHG - Atomic exchange
2661 Syntax: ``ATOMXCHG dst, resource, offset, src``
2663 Example: ``ATOMXCHG TEMP[0], BUFFER[0], TEMP[1], TEMP[2]``
2665 The following operation is performed atomically:
2669 dst_x = resource[offset]
2671 resource[offset] = src_x
2674 .. opcode:: ATOMCAS - Atomic compare-and-exchange
2676 Syntax: ``ATOMCAS dst, resource, offset, cmp, src``
2678 Example: ``ATOMCAS TEMP[0], BUFFER[0], TEMP[1], TEMP[2], TEMP[3]``
2680 The following operation is performed atomically:
2684 dst_x = resource[offset]
2686 resource[offset] = (dst_x == cmp_x ? src_x : dst_x)
2689 .. opcode:: ATOMAND - Atomic bitwise And
2691 Syntax: ``ATOMAND dst, resource, offset, src``
2693 Example: ``ATOMAND TEMP[0], BUFFER[0], TEMP[1], TEMP[2]``
2695 The following operation is performed atomically:
2699 dst_x = resource[offset]
2701 resource[offset] = dst_x \& src_x
2704 .. opcode:: ATOMOR - Atomic bitwise Or
2706 Syntax: ``ATOMOR dst, resource, offset, src``
2708 Example: ``ATOMOR TEMP[0], BUFFER[0], TEMP[1], TEMP[2]``
2710 The following operation is performed atomically:
2714 dst_x = resource[offset]
2716 resource[offset] = dst_x | src_x
2719 .. opcode:: ATOMXOR - Atomic bitwise Xor
2721 Syntax: ``ATOMXOR dst, resource, offset, src``
2723 Example: ``ATOMXOR TEMP[0], BUFFER[0], TEMP[1], TEMP[2]``
2725 The following operation is performed atomically:
2729 dst_x = resource[offset]
2731 resource[offset] = dst_x \oplus src_x
2734 .. opcode:: ATOMUMIN - Atomic unsigned minimum
2736 Syntax: ``ATOMUMIN dst, resource, offset, src``
2738 Example: ``ATOMUMIN TEMP[0], BUFFER[0], TEMP[1], TEMP[2]``
2740 The following operation is performed atomically:
2744 dst_x = resource[offset]
2746 resource[offset] = (dst_x < src_x ? dst_x : src_x)
2749 .. opcode:: ATOMUMAX - Atomic unsigned maximum
2751 Syntax: ``ATOMUMAX dst, resource, offset, src``
2753 Example: ``ATOMUMAX TEMP[0], BUFFER[0], TEMP[1], TEMP[2]``
2755 The following operation is performed atomically:
2759 dst_x = resource[offset]
2761 resource[offset] = (dst_x > src_x ? dst_x : src_x)
2764 .. opcode:: ATOMIMIN - Atomic signed minimum
2766 Syntax: ``ATOMIMIN dst, resource, offset, src``
2768 Example: ``ATOMIMIN TEMP[0], BUFFER[0], TEMP[1], TEMP[2]``
2770 The following operation is performed atomically:
2774 dst_x = resource[offset]
2776 resource[offset] = (dst_x < src_x ? dst_x : src_x)
2779 .. opcode:: ATOMIMAX - Atomic signed maximum
2781 Syntax: ``ATOMIMAX dst, resource, offset, src``
2783 Example: ``ATOMIMAX TEMP[0], BUFFER[0], TEMP[1], TEMP[2]``
2785 The following operation is performed atomically:
2789 dst_x = resource[offset]
2791 resource[offset] = (dst_x > src_x ? dst_x : src_x)
2794 .. _interlaneopcodes:
2799 These opcodes reduce the given value across the shader invocations
2800 running in the current SIMD group. Every thread in the subgroup will receive
2801 the same result. The BALLOT operations accept a single-channel argument that
2802 is treated as a boolean and produce a 64-bit value.
2804 .. opcode:: VOTE_ANY - Value is set in any of the active invocations
2806 Syntax: ``VOTE_ANY dst, value``
2808 Example: ``VOTE_ANY TEMP[0].x, TEMP[1].x``
2811 .. opcode:: VOTE_ALL - Value is set in all of the active invocations
2813 Syntax: ``VOTE_ALL dst, value``
2815 Example: ``VOTE_ALL TEMP[0].x, TEMP[1].x``
2818 .. opcode:: VOTE_EQ - Value is the same in all of the active invocations
2820 Syntax: ``VOTE_EQ dst, value``
2822 Example: ``VOTE_EQ TEMP[0].x, TEMP[1].x``
2825 .. opcode:: BALLOT - Lanemask of whether the value is set in each active
2828 Syntax: ``BALLOT dst, value``
2830 Example: ``BALLOT TEMP[0].xy, TEMP[1].x``
2832 When the argument is a constant true, this produces a bitmask of active
2833 invocations. In fragment shaders, this can include helper invocations
2834 (invocations whose outputs and writes to memory are discarded, but which
2835 are used to compute derivatives).
2838 .. opcode:: READ_FIRST - Broadcast the value from the first active
2839 invocation to all active lanes
2841 Syntax: ``READ_FIRST dst, value``
2843 Example: ``READ_FIRST TEMP[0], TEMP[1]``
2846 .. opcode:: READ_INVOC - Retrieve the value from the given invocation
2847 (need not be uniform)
2849 Syntax: ``READ_INVOC dst, value, invocation``
2851 Example: ``READ_INVOC TEMP[0].xy, TEMP[1].xy, TEMP[2].x``
2853 invocation.x controls the invocation number to read from for all channels.
2854 The invocation number must be the same across all active invocations in a
2855 sub-group; otherwise, the results are undefined.
2858 Explanation of symbols used
2859 ------------------------------
2866 :math:`|x|` Absolute value of `x`.
2868 :math:`\lceil x \rceil` Ceiling of `x`.
2870 clamp(x,y,z) Clamp x between y and z.
2871 (x < y) ? y : (x > z) ? z : x
2873 :math:`\lfloor x\rfloor` Floor of `x`.
2875 :math:`\log_2{x}` Logarithm of `x`, base 2.
2877 max(x,y) Maximum of x and y.
2880 min(x,y) Minimum of x and y.
2883 partialx(x) Derivative of x relative to fragment's X.
2885 partialy(x) Derivative of x relative to fragment's Y.
2887 pop() Pop from stack.
2889 :math:`x^y` `x` to the power `y`.
2891 push(x) Push x on stack.
2895 trunc(x) Truncate x, i.e. drop the fraction bits.
2902 discard Discard fragment.
2906 target Label of target instruction.
2917 Declares a register that is will be referenced as an operand in Instruction
2920 File field contains register file that is being declared and is one
2923 UsageMask field specifies which of the register components can be accessed
2924 and is one of TGSI_WRITEMASK.
2926 The Local flag specifies that a given value isn't intended for
2927 subroutine parameter passing and, as a result, the implementation
2928 isn't required to give any guarantees of it being preserved across
2929 subroutine boundaries. As it's merely a compiler hint, the
2930 implementation is free to ignore it.
2932 If Dimension flag is set to 1, a Declaration Dimension token follows.
2934 If Semantic flag is set to 1, a Declaration Semantic token follows.
2936 If Interpolate flag is set to 1, a Declaration Interpolate token follows.
2938 If file is TGSI_FILE_RESOURCE, a Declaration Resource token follows.
2940 If Array flag is set to 1, a Declaration Array token follows.
2943 ^^^^^^^^^^^^^^^^^^^^^^^^
2945 Declarations can optional have an ArrayID attribute which can be referred by
2946 indirect addressing operands. An ArrayID of zero is reserved and treated as
2947 if no ArrayID is specified.
2949 If an indirect addressing operand refers to a specific declaration by using
2950 an ArrayID only the registers in this declaration are guaranteed to be
2951 accessed, accessing any register outside this declaration results in undefined
2952 behavior. Note that for compatibility the effective index is zero-based and
2953 not relative to the specified declaration
2955 If no ArrayID is specified with an indirect addressing operand the whole
2956 register file might be accessed by this operand. This is strongly discouraged
2957 and will prevent packing of scalar/vec2 arrays and effective alias analysis.
2958 This is only legal for TEMP and CONST register files.
2960 Declaration Semantic
2961 ^^^^^^^^^^^^^^^^^^^^^^^^
2963 Vertex and fragment shader input and output registers may be labeled
2964 with semantic information consisting of a name and index.
2966 Follows Declaration token if Semantic bit is set.
2968 Since its purpose is to link a shader with other stages of the pipeline,
2969 it is valid to follow only those Declaration tokens that declare a register
2970 either in INPUT or OUTPUT file.
2972 SemanticName field contains the semantic name of the register being declared.
2973 There is no default value.
2975 SemanticIndex is an optional subscript that can be used to distinguish
2976 different register declarations with the same semantic name. The default value
2979 The meanings of the individual semantic names are explained in the following
2982 TGSI_SEMANTIC_POSITION
2983 """"""""""""""""""""""
2985 For vertex shaders, TGSI_SEMANTIC_POSITION indicates the vertex shader
2986 output register which contains the homogeneous vertex position in the clip
2987 space coordinate system. After clipping, the X, Y and Z components of the
2988 vertex will be divided by the W value to get normalized device coordinates.
2990 For fragment shaders, TGSI_SEMANTIC_POSITION is used to indicate that
2991 fragment shader input (or system value, depending on which one is
2992 supported by the driver) contains the fragment's window position. The X
2993 component starts at zero and always increases from left to right.
2994 The Y component starts at zero and always increases but Y=0 may either
2995 indicate the top of the window or the bottom depending on the fragment
2996 coordinate origin convention (see TGSI_PROPERTY_FS_COORD_ORIGIN).
2997 The Z coordinate ranges from 0 to 1 to represent depth from the front
2998 to the back of the Z buffer. The W component contains the interpolated
2999 reciprocal of the vertex position W component (corresponding to gl_Fragcoord,
3000 but unlike d3d10 which interpolates the same 1/w but then gives back
3001 the reciprocal of the interpolated value).
3003 Fragment shaders may also declare an output register with
3004 TGSI_SEMANTIC_POSITION. Only the Z component is writable. This allows
3005 the fragment shader to change the fragment's Z position.
3012 For vertex shader outputs or fragment shader inputs/outputs, this
3013 label indicates that the register contains an R,G,B,A color.
3015 Several shader inputs/outputs may contain colors so the semantic index
3016 is used to distinguish them. For example, color[0] may be the diffuse
3017 color while color[1] may be the specular color.
3019 This label is needed so that the flat/smooth shading can be applied
3020 to the right interpolants during rasterization.
3024 TGSI_SEMANTIC_BCOLOR
3025 """"""""""""""""""""
3027 Back-facing colors are only used for back-facing polygons, and are only valid
3028 in vertex shader outputs. After rasterization, all polygons are front-facing
3029 and COLOR and BCOLOR end up occupying the same slots in the fragment shader,
3030 so all BCOLORs effectively become regular COLORs in the fragment shader.
3036 Vertex shader inputs and outputs and fragment shader inputs may be
3037 labeled with TGSI_SEMANTIC_FOG to indicate that the register contains
3038 a fog coordinate. Typically, the fragment shader will use the fog coordinate
3039 to compute a fog blend factor which is used to blend the normal fragment color
3040 with a constant fog color. But fog coord really is just an ordinary vec4
3041 register like regular semantics.
3047 Vertex shader input and output registers may be labeled with
3048 TGIS_SEMANTIC_PSIZE to indicate that the register contains a point size
3049 in the form (S, 0, 0, 1). The point size controls the width or diameter
3050 of points for rasterization. This label cannot be used in fragment
3053 When using this semantic, be sure to set the appropriate state in the
3054 :ref:`rasterizer` first.
3057 TGSI_SEMANTIC_TEXCOORD
3058 """"""""""""""""""""""
3060 Only available if PIPE_CAP_TGSI_TEXCOORD is exposed !
3062 Vertex shader outputs and fragment shader inputs may be labeled with
3063 this semantic to make them replaceable by sprite coordinates via the
3064 sprite_coord_enable state in the :ref:`rasterizer`.
3065 The semantic index permitted with this semantic is limited to <= 7.
3067 If the driver does not support TEXCOORD, sprite coordinate replacement
3068 applies to inputs with the GENERIC semantic instead.
3070 The intended use case for this semantic is gl_TexCoord.
3073 TGSI_SEMANTIC_PCOORD
3074 """"""""""""""""""""
3076 Only available if PIPE_CAP_TGSI_TEXCOORD is exposed !
3078 Fragment shader inputs may be labeled with TGSI_SEMANTIC_PCOORD to indicate
3079 that the register contains sprite coordinates in the form (x, y, 0, 1), if
3080 the current primitive is a point and point sprites are enabled. Otherwise,
3081 the contents of the register are undefined.
3083 The intended use case for this semantic is gl_PointCoord.
3086 TGSI_SEMANTIC_GENERIC
3087 """""""""""""""""""""
3089 All vertex/fragment shader inputs/outputs not labeled with any other
3090 semantic label can be considered to be generic attributes. Typical
3091 uses of generic inputs/outputs are texcoords and user-defined values.
3094 TGSI_SEMANTIC_NORMAL
3095 """"""""""""""""""""
3097 Indicates that a vertex shader input is a normal vector. This is
3098 typically only used for legacy graphics APIs.
3104 This label applies to fragment shader inputs (or system values,
3105 depending on which one is supported by the driver) and indicates that
3106 the register contains front/back-face information.
3108 If it is an input, it will be a floating-point vector in the form (F, 0, 0, 1),
3109 where F will be positive when the fragment belongs to a front-facing polygon,
3110 and negative when the fragment belongs to a back-facing polygon.
3112 If it is a system value, it will be an integer vector in the form (F, 0, 0, 1),
3113 where F is 0xffffffff when the fragment belongs to a front-facing polygon and
3114 0 when the fragment belongs to a back-facing polygon.
3117 TGSI_SEMANTIC_EDGEFLAG
3118 """"""""""""""""""""""
3120 For vertex shaders, this sematic label indicates that an input or
3121 output is a boolean edge flag. The register layout is [F, x, x, x]
3122 where F is 0.0 or 1.0 and x = don't care. Normally, the vertex shader
3123 simply copies the edge flag input to the edgeflag output.
3125 Edge flags are used to control which lines or points are actually
3126 drawn when the polygon mode converts triangles/quads/polygons into
3130 TGSI_SEMANTIC_STENCIL
3131 """""""""""""""""""""
3133 For fragment shaders, this semantic label indicates that an output
3134 is a writable stencil reference value. Only the Y component is writable.
3135 This allows the fragment shader to change the fragments stencilref value.
3138 TGSI_SEMANTIC_VIEWPORT_INDEX
3139 """"""""""""""""""""""""""""
3141 For geometry shaders, this semantic label indicates that an output
3142 contains the index of the viewport (and scissor) to use.
3143 This is an integer value, and only the X component is used.
3145 If PIPE_CAP_TGSI_VS_LAYER_VIEWPORT or PIPE_CAP_TGSI_TES_LAYER_VIEWPORT is
3146 supported, then this semantic label can also be used in vertex or
3147 tessellation evaluation shaders, respectively. Only the value written in the
3148 last vertex processing stage is used.
3154 For geometry shaders, this semantic label indicates that an output
3155 contains the layer value to use for the color and depth/stencil surfaces.
3156 This is an integer value, and only the X component is used.
3157 (Also known as rendertarget array index.)
3159 If PIPE_CAP_TGSI_VS_LAYER_VIEWPORT or PIPE_CAP_TGSI_TES_LAYER_VIEWPORT is
3160 supported, then this semantic label can also be used in vertex or
3161 tessellation evaluation shaders, respectively. Only the value written in the
3162 last vertex processing stage is used.
3165 TGSI_SEMANTIC_CULLDIST
3166 """"""""""""""""""""""
3168 Used as distance to plane for performing application-defined culling
3169 of individual primitives against a plane. When components of vertex
3170 elements are given this label, these values are assumed to be a
3171 float32 signed distance to a plane. Primitives will be completely
3172 discarded if the plane distance for all of the vertices in the
3173 primitive are < 0. If a vertex has a cull distance of NaN, that
3174 vertex counts as "out" (as if its < 0);
3175 The limits on both clip and cull distances are bound
3176 by the PIPE_MAX_CLIP_OR_CULL_DISTANCE_COUNT define which defines
3177 the maximum number of components that can be used to hold the
3178 distances and by the PIPE_MAX_CLIP_OR_CULL_DISTANCE_ELEMENT_COUNT
3179 which specifies the maximum number of registers which can be
3180 annotated with those semantics.
3183 TGSI_SEMANTIC_CLIPDIST
3184 """"""""""""""""""""""
3186 Note this covers clipping and culling distances.
3188 When components of vertex elements are identified this way, these
3189 values are each assumed to be a float32 signed distance to a plane.
3192 Primitive setup only invokes rasterization on pixels for which
3193 the interpolated plane distances are >= 0.
3196 Primitives will be completely discarded if the plane distance
3197 for all of the vertices in the primitive are < 0.
3198 If a vertex has a cull distance of NaN, that vertex counts as "out"
3201 Multiple clip/cull planes can be implemented simultaneously, by
3202 annotating multiple components of one or more vertex elements with
3203 the above specified semantic.
3204 The limits on both clip and cull distances are bound
3205 by the PIPE_MAX_CLIP_OR_CULL_DISTANCE_COUNT define which defines
3206 the maximum number of components that can be used to hold the
3207 distances and by the PIPE_MAX_CLIP_OR_CULL_DISTANCE_ELEMENT_COUNT
3208 which specifies the maximum number of registers which can be
3209 annotated with those semantics.
3210 The properties NUM_CLIPDIST_ENABLED and NUM_CULLDIST_ENABLED
3211 are used to divide up the 2 x vec4 space between clipping and culling.
3213 TGSI_SEMANTIC_SAMPLEID
3214 """"""""""""""""""""""
3216 For fragment shaders, this semantic label indicates that a system value
3217 contains the current sample id (i.e. gl_SampleID) as an unsigned int.
3218 Only the X component is used. If per-sample shading is not enabled,
3219 the result is (0, undef, undef, undef).
3221 Note that if the fragment shader uses this system value, the fragment
3222 shader is automatically executed at per sample frequency.
3224 TGSI_SEMANTIC_SAMPLEPOS
3225 """""""""""""""""""""""
3227 For fragment shaders, this semantic label indicates that a system
3228 value contains the current sample's position as float4(x, y, undef, undef)
3229 in the render target (i.e. gl_SamplePosition) when per-fragment shading
3230 is in effect. Position values are in the range [0, 1] where 0.5 is
3231 the center of the fragment.
3233 Note that if the fragment shader uses this system value, the fragment
3234 shader is automatically executed at per sample frequency.
3236 TGSI_SEMANTIC_SAMPLEMASK
3237 """"""""""""""""""""""""
3239 For fragment shaders, this semantic label can be applied to either a
3240 shader system value input or output.
3242 For a system value, the sample mask indicates the set of samples covered by
3243 the current primitive. If MSAA is not enabled, the value is (1, 0, 0, 0).
3245 For an output, the sample mask is used to disable further sample processing.
3247 For both, the register type is uint[4] but only the X component is used
3248 (i.e. gl_SampleMask[0]). Each bit corresponds to one sample position (up
3249 to 32x MSAA is supported).
3251 TGSI_SEMANTIC_INVOCATIONID
3252 """"""""""""""""""""""""""
3254 For geometry shaders, this semantic label indicates that a system value
3255 contains the current invocation id (i.e. gl_InvocationID).
3256 This is an integer value, and only the X component is used.
3258 TGSI_SEMANTIC_INSTANCEID
3259 """"""""""""""""""""""""
3261 For vertex shaders, this semantic label indicates that a system value contains
3262 the current instance id (i.e. gl_InstanceID). It does not include the base
3263 instance. This is an integer value, and only the X component is used.
3265 TGSI_SEMANTIC_VERTEXID
3266 """"""""""""""""""""""
3268 For vertex shaders, this semantic label indicates that a system value contains
3269 the current vertex id (i.e. gl_VertexID). It does (unlike in d3d10) include the
3270 base vertex. This is an integer value, and only the X component is used.
3272 TGSI_SEMANTIC_VERTEXID_NOBASE
3273 """""""""""""""""""""""""""""""
3275 For vertex shaders, this semantic label indicates that a system value contains
3276 the current vertex id without including the base vertex (this corresponds to
3277 d3d10 vertex id, so TGSI_SEMANTIC_VERTEXID_NOBASE + TGSI_SEMANTIC_BASEVERTEX
3278 == TGSI_SEMANTIC_VERTEXID). This is an integer value, and only the X component
3281 TGSI_SEMANTIC_BASEVERTEX
3282 """"""""""""""""""""""""
3284 For vertex shaders, this semantic label indicates that a system value contains
3285 the base vertex (i.e. gl_BaseVertex). Note that for non-indexed draw calls,
3286 this contains the first (or start) value instead.
3287 This is an integer value, and only the X component is used.
3289 TGSI_SEMANTIC_PRIMID
3290 """"""""""""""""""""
3292 For geometry and fragment shaders, this semantic label indicates the value
3293 contains the primitive id (i.e. gl_PrimitiveID). This is an integer value,
3294 and only the X component is used.
3295 FIXME: This right now can be either a ordinary input or a system value...
3301 For tessellation evaluation/control shaders, this semantic label indicates a
3302 generic per-patch attribute. Such semantics will not implicitly be per-vertex
3305 TGSI_SEMANTIC_TESSCOORD
3306 """""""""""""""""""""""
3308 For tessellation evaluation shaders, this semantic label indicates the
3309 coordinates of the vertex being processed. This is available in XYZ; W is
3312 TGSI_SEMANTIC_TESSOUTER
3313 """""""""""""""""""""""
3315 For tessellation evaluation/control shaders, this semantic label indicates the
3316 outer tessellation levels of the patch. Isoline tessellation will only have XY
3317 defined, triangle will have XYZ and quads will have XYZW defined. This
3318 corresponds to gl_TessLevelOuter.
3320 TGSI_SEMANTIC_TESSINNER
3321 """""""""""""""""""""""
3323 For tessellation evaluation/control shaders, this semantic label indicates the
3324 inner tessellation levels of the patch. The X value is only defined for
3325 triangle tessellation, while quads will have XY defined. This is entirely
3326 undefined for isoline tessellation.
3328 TGSI_SEMANTIC_VERTICESIN
3329 """"""""""""""""""""""""
3331 For tessellation evaluation/control shaders, this semantic label indicates the
3332 number of vertices provided in the input patch. Only the X value is defined.
3334 TGSI_SEMANTIC_HELPER_INVOCATION
3335 """""""""""""""""""""""""""""""
3337 For fragment shaders, this semantic indicates whether the current
3338 invocation is covered or not. Helper invocations are created in order
3339 to properly compute derivatives, however it may be desirable to skip
3340 some of the logic in those cases. See ``gl_HelperInvocation`` documentation.
3342 TGSI_SEMANTIC_BASEINSTANCE
3343 """"""""""""""""""""""""""
3345 For vertex shaders, the base instance argument supplied for this
3346 draw. This is an integer value, and only the X component is used.
3348 TGSI_SEMANTIC_DRAWID
3349 """"""""""""""""""""
3351 For vertex shaders, the zero-based index of the current draw in a
3352 ``glMultiDraw*`` invocation. This is an integer value, and only the X
3356 TGSI_SEMANTIC_WORK_DIM
3357 """"""""""""""""""""""
3359 For compute shaders started via opencl this retrieves the work_dim
3360 parameter to the clEnqueueNDRangeKernel call with which the shader
3364 TGSI_SEMANTIC_GRID_SIZE
3365 """""""""""""""""""""""
3367 For compute shaders, this semantic indicates the maximum (x, y, z) dimensions
3368 of a grid of thread blocks.
3371 TGSI_SEMANTIC_BLOCK_ID
3372 """"""""""""""""""""""
3374 For compute shaders, this semantic indicates the (x, y, z) coordinates of the
3375 current block inside of the grid.
3378 TGSI_SEMANTIC_BLOCK_SIZE
3379 """"""""""""""""""""""""
3381 For compute shaders, this semantic indicates the maximum (x, y, z) dimensions
3382 of a block in threads.
3385 TGSI_SEMANTIC_THREAD_ID
3386 """""""""""""""""""""""
3388 For compute shaders, this semantic indicates the (x, y, z) coordinates of the
3389 current thread inside of the block.
3392 TGSI_SEMANTIC_SUBGROUP_SIZE
3393 """""""""""""""""""""""""""
3395 This semantic indicates the subgroup size for the current invocation. This is
3396 an integer of at most 64, as it indicates the width of lanemasks. It does not
3397 depend on the number of invocations that are active.
3400 TGSI_SEMANTIC_SUBGROUP_INVOCATION
3401 """""""""""""""""""""""""""""""""
3403 The index of the current invocation within its subgroup.
3406 TGSI_SEMANTIC_SUBGROUP_EQ_MASK
3407 """"""""""""""""""""""""""""""
3409 A bit mask of ``bit index == TGSI_SEMANTIC_SUBGROUP_INVOCATION``, i.e.
3410 ``1 << subgroup_invocation`` in arbitrary precision arithmetic.
3413 TGSI_SEMANTIC_SUBGROUP_GE_MASK
3414 """"""""""""""""""""""""""""""
3416 A bit mask of ``bit index >= TGSI_SEMANTIC_SUBGROUP_INVOCATION``, i.e.
3417 ``((1 << (subgroup_size - subgroup_invocation)) - 1) << subgroup_invocation``
3418 in arbitrary precision arithmetic.
3421 TGSI_SEMANTIC_SUBGROUP_GT_MASK
3422 """"""""""""""""""""""""""""""
3424 A bit mask of ``bit index > TGSI_SEMANTIC_SUBGROUP_INVOCATION``, i.e.
3425 ``((1 << (subgroup_size - subgroup_invocation - 1)) - 1) << (subgroup_invocation + 1)``
3426 in arbitrary precision arithmetic.
3429 TGSI_SEMANTIC_SUBGROUP_LE_MASK
3430 """"""""""""""""""""""""""""""
3432 A bit mask of ``bit index <= TGSI_SEMANTIC_SUBGROUP_INVOCATION``, i.e.
3433 ``(1 << (subgroup_invocation + 1)) - 1`` in arbitrary precision arithmetic.
3436 TGSI_SEMANTIC_SUBGROUP_LT_MASK
3437 """"""""""""""""""""""""""""""
3439 A bit mask of ``bit index > TGSI_SEMANTIC_SUBGROUP_INVOCATION``, i.e.
3440 ``(1 << subgroup_invocation) - 1`` in arbitrary precision arithmetic.
3443 Declaration Interpolate
3444 ^^^^^^^^^^^^^^^^^^^^^^^
3446 This token is only valid for fragment shader INPUT declarations.
3448 The Interpolate field specifes the way input is being interpolated by
3449 the rasteriser and is one of TGSI_INTERPOLATE_*.
3451 The Location field specifies the location inside the pixel that the
3452 interpolation should be done at, one of ``TGSI_INTERPOLATE_LOC_*``. Note that
3453 when per-sample shading is enabled, the implementation may choose to
3454 interpolate at the sample irrespective of the Location field.
3456 The CylindricalWrap bitfield specifies which register components
3457 should be subject to cylindrical wrapping when interpolating by the
3458 rasteriser. If TGSI_CYLINDRICAL_WRAP_X is set to 1, the X component
3459 should be interpolated according to cylindrical wrapping rules.
3462 Declaration Sampler View
3463 ^^^^^^^^^^^^^^^^^^^^^^^^
3465 Follows Declaration token if file is TGSI_FILE_SAMPLER_VIEW.
3467 DCL SVIEW[#], resource, type(s)
3469 Declares a shader input sampler view and assigns it to a SVIEW[#]
3472 resource can be one of BUFFER, 1D, 2D, 3D, 1DArray and 2DArray.
3474 type must be 1 or 4 entries (if specifying on a per-component
3475 level) out of UNORM, SNORM, SINT, UINT and FLOAT.
3477 For TEX\* style texture sample opcodes (as opposed to SAMPLE\* opcodes
3478 which take an explicit SVIEW[#] source register), there may be optionally
3479 SVIEW[#] declarations. In this case, the SVIEW index is implied by the
3480 SAMP index, and there must be a corresponding SVIEW[#] declaration for
3481 each SAMP[#] declaration. Drivers are free to ignore this if they wish.
3482 But note in particular that some drivers need to know the sampler type
3483 (float/int/unsigned) in order to generate the correct code, so cases
3484 where integer textures are sampled, SVIEW[#] declarations should be
3487 NOTE: It is NOT legal to mix SAMPLE\* style opcodes and TEX\* opcodes
3490 Declaration Resource
3491 ^^^^^^^^^^^^^^^^^^^^
3493 Follows Declaration token if file is TGSI_FILE_RESOURCE.
3495 DCL RES[#], resource [, WR] [, RAW]
3497 Declares a shader input resource and assigns it to a RES[#]
3500 resource can be one of BUFFER, 1D, 2D, 3D, CUBE, 1DArray and
3503 If the RAW keyword is not specified, the texture data will be
3504 subject to conversion, swizzling and scaling as required to yield
3505 the specified data type from the physical data format of the bound
3508 If the RAW keyword is specified, no channel conversion will be
3509 performed: the values read for each of the channels (X,Y,Z,W) will
3510 correspond to consecutive words in the same order and format
3511 they're found in memory. No element-to-address conversion will be
3512 performed either: the value of the provided X coordinate will be
3513 interpreted in byte units instead of texel units. The result of
3514 accessing a misaligned address is undefined.
3516 Usage of the STORE opcode is only allowed if the WR (writable) flag
3521 ^^^^^^^^^^^^^^^^^^^^^^^^
3523 Properties are general directives that apply to the whole TGSI program.
3528 Specifies the fragment shader TGSI_SEMANTIC_POSITION coordinate origin.
3529 The default value is UPPER_LEFT.
3531 If UPPER_LEFT, the position will be (0,0) at the upper left corner and
3532 increase downward and rightward.
3533 If LOWER_LEFT, the position will be (0,0) at the lower left corner and
3534 increase upward and rightward.
3536 OpenGL defaults to LOWER_LEFT, and is configurable with the
3537 GL_ARB_fragment_coord_conventions extension.
3539 DirectX 9/10 use UPPER_LEFT.
3541 FS_COORD_PIXEL_CENTER
3542 """""""""""""""""""""
3544 Specifies the fragment shader TGSI_SEMANTIC_POSITION pixel center convention.
3545 The default value is HALF_INTEGER.
3547 If HALF_INTEGER, the fractionary part of the position will be 0.5
3548 If INTEGER, the fractionary part of the position will be 0.0
3550 Note that this does not affect the set of fragments generated by
3551 rasterization, which is instead controlled by half_pixel_center in the
3554 OpenGL defaults to HALF_INTEGER, and is configurable with the
3555 GL_ARB_fragment_coord_conventions extension.
3557 DirectX 9 uses INTEGER.
3558 DirectX 10 uses HALF_INTEGER.
3560 FS_COLOR0_WRITES_ALL_CBUFS
3561 """"""""""""""""""""""""""
3562 Specifies that writes to the fragment shader color 0 are replicated to all
3563 bound cbufs. This facilitates OpenGL's fragColor output vs fragData[0] where
3564 fragData is directed to a single color buffer, but fragColor is broadcast.
3567 """"""""""""""""""""""""""
3568 If this property is set on the program bound to the shader stage before the
3569 fragment shader, user clip planes should have no effect (be disabled) even if
3570 that shader does not write to any clip distance outputs and the rasterizer's
3571 clip_plane_enable is non-zero.
3572 This property is only supported by drivers that also support shader clip
3574 This is useful for APIs that don't have UCPs and where clip distances written
3575 by a shader cannot be disabled.
3580 Specifies the number of times a geometry shader should be executed for each
3581 input primitive. Each invocation will have a different
3582 TGSI_SEMANTIC_INVOCATIONID system value set. If not specified, assumed to
3585 VS_WINDOW_SPACE_POSITION
3586 """"""""""""""""""""""""""
3587 If this property is set on the vertex shader, the TGSI_SEMANTIC_POSITION output
3588 is assumed to contain window space coordinates.
3589 Division of X,Y,Z by W and the viewport transformation are disabled, and 1/W is
3590 directly taken from the 4-th component of the shader output.
3591 Naturally, clipping is not performed on window coordinates either.
3592 The effect of this property is undefined if a geometry or tessellation shader
3598 The number of vertices written by the tessellation control shader. This
3599 effectively defines the patch input size of the tessellation evaluation shader
3605 This sets the tessellation primitive mode, one of ``PIPE_PRIM_TRIANGLES``,
3606 ``PIPE_PRIM_QUADS``, or ``PIPE_PRIM_LINES``. (Unlike in GL, there is no
3607 separate isolines settings, the regular lines is assumed to mean isolines.)
3612 This sets the spacing mode of the tessellation generator, one of
3613 ``PIPE_TESS_SPACING_*``.
3618 This sets the vertex order to be clockwise if the value is 1, or
3619 counter-clockwise if set to 0.
3624 If set to a non-zero value, this turns on point mode for the tessellator,
3625 which means that points will be generated instead of primitives.
3627 NUM_CLIPDIST_ENABLED
3628 """"""""""""""""""""
3630 How many clip distance scalar outputs are enabled.
3632 NUM_CULLDIST_ENABLED
3633 """"""""""""""""""""
3635 How many cull distance scalar outputs are enabled.
3637 FS_EARLY_DEPTH_STENCIL
3638 """"""""""""""""""""""
3640 Whether depth test, stencil test, and occlusion query should run before
3641 the fragment shader (regardless of fragment shader side effects). Corresponds
3642 to GLSL early_fragment_tests.
3647 Which shader stage will MOST LIKELY follow after this shader when the shader
3648 is bound. This is only a hint to the driver and doesn't have to be precise.
3649 Only set for VS and TES.
3651 CS_FIXED_BLOCK_WIDTH / HEIGHT / DEPTH
3652 """""""""""""""""""""""""""""""""""""
3654 Threads per block in each dimension, if known at compile time. If the block size
3655 is known all three should be at least 1. If it is unknown they should all be set
3661 The MUL TGSI operation (FP32 multiplication) will return 0 if either
3662 of the operands are equal to 0. That means that 0 * Inf = 0. This
3663 should be set the same way for an entire pipeline. Note that this
3664 applies not only to the literal MUL TGSI opcode, but all FP32
3665 multiplications implied by other operations, such as MAD, FMA, DP2,
3666 DP3, DP4, DST, LOG, LRP, XPD, and possibly others. If there is a
3667 mismatch between shaders, then it is unspecified whether this behavior
3670 FS_POST_DEPTH_COVERAGE
3671 """"""""""""""""""""""
3673 When enabled, the input for TGSI_SEMANTIC_SAMPLEMASK will exclude samples
3674 that have failed the depth/stencil tests. This is only valid when
3675 FS_EARLY_DEPTH_STENCIL is also specified.
3678 Texture Sampling and Texture Formats
3679 ------------------------------------
3681 This table shows how texture image components are returned as (x,y,z,w) tuples
3682 by TGSI texture instructions, such as :opcode:`TEX`, :opcode:`TXD`, and
3683 :opcode:`TXP`. For reference, OpenGL and Direct3D conventions are shown as
3686 +--------------------+--------------+--------------------+--------------+
3687 | Texture Components | Gallium | OpenGL | Direct3D 9 |
3688 +====================+==============+====================+==============+
3689 | R | (r, 0, 0, 1) | (r, 0, 0, 1) | (r, 1, 1, 1) |
3690 +--------------------+--------------+--------------------+--------------+
3691 | RG | (r, g, 0, 1) | (r, g, 0, 1) | (r, g, 1, 1) |
3692 +--------------------+--------------+--------------------+--------------+
3693 | RGB | (r, g, b, 1) | (r, g, b, 1) | (r, g, b, 1) |
3694 +--------------------+--------------+--------------------+--------------+
3695 | RGBA | (r, g, b, a) | (r, g, b, a) | (r, g, b, a) |
3696 +--------------------+--------------+--------------------+--------------+
3697 | A | (0, 0, 0, a) | (0, 0, 0, a) | (0, 0, 0, a) |
3698 +--------------------+--------------+--------------------+--------------+
3699 | L | (l, l, l, 1) | (l, l, l, 1) | (l, l, l, 1) |
3700 +--------------------+--------------+--------------------+--------------+
3701 | LA | (l, l, l, a) | (l, l, l, a) | (l, l, l, a) |
3702 +--------------------+--------------+--------------------+--------------+
3703 | I | (i, i, i, i) | (i, i, i, i) | N/A |
3704 +--------------------+--------------+--------------------+--------------+
3705 | UV | XXX TBD | (0, 0, 0, 1) | (u, v, 1, 1) |
3706 | | | [#envmap-bumpmap]_ | |
3707 +--------------------+--------------+--------------------+--------------+
3708 | Z | XXX TBD | (z, z, z, 1) | (0, z, 0, 1) |
3709 | | | [#depth-tex-mode]_ | |
3710 +--------------------+--------------+--------------------+--------------+
3711 | S | (s, s, s, s) | unknown | unknown |
3712 +--------------------+--------------+--------------------+--------------+
3714 .. [#envmap-bumpmap] http://www.opengl.org/registry/specs/ATI/envmap_bumpmap.txt
3715 .. [#depth-tex-mode] the default is (z, z, z, 1) but may also be (0, 0, 0, z)
3716 or (z, z, z, z) depending on the value of GL_DEPTH_TEXTURE_MODE.