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:`Double Opcodes`.
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.
30 ^^^^^^^^^^^^^^^^^^^^^^^^^
32 These opcodes are guaranteed to be available regardless of the driver being
35 .. opcode:: ARL - Address Register Load
39 dst.x = \lfloor src.x\rfloor
41 dst.y = \lfloor src.y\rfloor
43 dst.z = \lfloor src.z\rfloor
45 dst.w = \lfloor src.w\rfloor
48 .. opcode:: MOV - Move
61 .. opcode:: LIT - Light Coefficients
69 dst.z = (src.x > 0) ? max(src.y, 0)^{clamp(src.w, -128, 128))} : 0
74 .. opcode:: RCP - Reciprocal
76 This instruction replicates its result.
83 .. opcode:: RSQ - Reciprocal Square Root
85 This instruction replicates its result.
89 dst = \frac{1}{\sqrt{|src.x|}}
92 .. opcode:: EXP - Approximate Exponential Base 2
96 dst.x = 2^{\lfloor src.x\rfloor}
98 dst.y = src.x - \lfloor src.x\rfloor
105 .. opcode:: LOG - Approximate Logarithm Base 2
109 dst.x = \lfloor\log_2{|src.x|}\rfloor
111 dst.y = \frac{|src.x|}{2^{\lfloor\log_2{|src.x|}\rfloor}}
113 dst.z = \log_2{|src.x|}
118 .. opcode:: MUL - Multiply
122 dst.x = src0.x \times src1.x
124 dst.y = src0.y \times src1.y
126 dst.z = src0.z \times src1.z
128 dst.w = src0.w \times src1.w
131 .. opcode:: ADD - Add
135 dst.x = src0.x + src1.x
137 dst.y = src0.y + src1.y
139 dst.z = src0.z + src1.z
141 dst.w = src0.w + src1.w
144 .. opcode:: DP3 - 3-component Dot Product
146 This instruction replicates its result.
150 dst = src0.x \times src1.x + src0.y \times src1.y + src0.z \times src1.z
153 .. opcode:: DP4 - 4-component Dot Product
155 This instruction replicates its result.
159 dst = src0.x \times src1.x + src0.y \times src1.y + src0.z \times src1.z + src0.w \times src1.w
162 .. opcode:: DST - Distance Vector
168 dst.y = src0.y \times src1.y
175 .. opcode:: MIN - Minimum
179 dst.x = min(src0.x, src1.x)
181 dst.y = min(src0.y, src1.y)
183 dst.z = min(src0.z, src1.z)
185 dst.w = min(src0.w, src1.w)
188 .. opcode:: MAX - Maximum
192 dst.x = max(src0.x, src1.x)
194 dst.y = max(src0.y, src1.y)
196 dst.z = max(src0.z, src1.z)
198 dst.w = max(src0.w, src1.w)
201 .. opcode:: SLT - Set On Less Than
205 dst.x = (src0.x < src1.x) ? 1 : 0
207 dst.y = (src0.y < src1.y) ? 1 : 0
209 dst.z = (src0.z < src1.z) ? 1 : 0
211 dst.w = (src0.w < src1.w) ? 1 : 0
214 .. opcode:: SGE - Set On Greater Equal Than
218 dst.x = (src0.x >= src1.x) ? 1 : 0
220 dst.y = (src0.y >= src1.y) ? 1 : 0
222 dst.z = (src0.z >= src1.z) ? 1 : 0
224 dst.w = (src0.w >= src1.w) ? 1 : 0
227 .. opcode:: MAD - Multiply And Add
231 dst.x = src0.x \times src1.x + src2.x
233 dst.y = src0.y \times src1.y + src2.y
235 dst.z = src0.z \times src1.z + src2.z
237 dst.w = src0.w \times src1.w + src2.w
240 .. opcode:: SUB - Subtract
244 dst.x = src0.x - src1.x
246 dst.y = src0.y - src1.y
248 dst.z = src0.z - src1.z
250 dst.w = src0.w - src1.w
253 .. opcode:: LRP - Linear Interpolate
257 dst.x = src0.x \times src1.x + (1 - src0.x) \times src2.x
259 dst.y = src0.y \times src1.y + (1 - src0.y) \times src2.y
261 dst.z = src0.z \times src1.z + (1 - src0.z) \times src2.z
263 dst.w = src0.w \times src1.w + (1 - src0.w) \times src2.w
266 .. opcode:: CND - Condition
270 dst.x = (src2.x > 0.5) ? src0.x : src1.x
272 dst.y = (src2.y > 0.5) ? src0.y : src1.y
274 dst.z = (src2.z > 0.5) ? src0.z : src1.z
276 dst.w = (src2.w > 0.5) ? src0.w : src1.w
279 .. opcode:: DP2A - 2-component Dot Product And Add
283 dst.x = src0.x \times src1.x + src0.y \times src1.y + src2.x
285 dst.y = src0.x \times src1.x + src0.y \times src1.y + src2.x
287 dst.z = src0.x \times src1.x + src0.y \times src1.y + src2.x
289 dst.w = src0.x \times src1.x + src0.y \times src1.y + src2.x
292 .. opcode:: FRC - Fraction
296 dst.x = src.x - \lfloor src.x\rfloor
298 dst.y = src.y - \lfloor src.y\rfloor
300 dst.z = src.z - \lfloor src.z\rfloor
302 dst.w = src.w - \lfloor src.w\rfloor
305 .. opcode:: CLAMP - Clamp
309 dst.x = clamp(src0.x, src1.x, src2.x)
311 dst.y = clamp(src0.y, src1.y, src2.y)
313 dst.z = clamp(src0.z, src1.z, src2.z)
315 dst.w = clamp(src0.w, src1.w, src2.w)
318 .. opcode:: FLR - Floor
320 This is identical to :opcode:`ARL`.
324 dst.x = \lfloor src.x\rfloor
326 dst.y = \lfloor src.y\rfloor
328 dst.z = \lfloor src.z\rfloor
330 dst.w = \lfloor src.w\rfloor
333 .. opcode:: ROUND - Round
346 .. opcode:: EX2 - Exponential Base 2
348 This instruction replicates its result.
355 .. opcode:: LG2 - Logarithm Base 2
357 This instruction replicates its result.
364 .. opcode:: POW - Power
366 This instruction replicates its result.
370 dst = src0.x^{src1.x}
372 .. opcode:: XPD - Cross Product
376 dst.x = src0.y \times src1.z - src1.y \times src0.z
378 dst.y = src0.z \times src1.x - src1.z \times src0.x
380 dst.z = src0.x \times src1.y - src1.x \times src0.y
385 .. opcode:: ABS - Absolute
398 .. opcode:: RCC - Reciprocal Clamped
400 This instruction replicates its result.
402 XXX cleanup on aisle three
406 dst = (1 / src.x) > 0 ? clamp(1 / src.x, 5.42101e-020, 1.884467e+019) : clamp(1 / src.x, -1.884467e+019, -5.42101e-020)
409 .. opcode:: DPH - Homogeneous Dot Product
411 This instruction replicates its result.
415 dst = src0.x \times src1.x + src0.y \times src1.y + src0.z \times src1.z + src1.w
418 .. opcode:: COS - Cosine
420 This instruction replicates its result.
427 .. opcode:: DDX - Derivative Relative To X
431 dst.x = partialx(src.x)
433 dst.y = partialx(src.y)
435 dst.z = partialx(src.z)
437 dst.w = partialx(src.w)
440 .. opcode:: DDY - Derivative Relative To Y
444 dst.x = partialy(src.x)
446 dst.y = partialy(src.y)
448 dst.z = partialy(src.z)
450 dst.w = partialy(src.w)
453 .. opcode:: KILP - Predicated Discard
458 .. opcode:: PK2H - Pack Two 16-bit Floats
463 .. opcode:: PK2US - Pack Two Unsigned 16-bit Scalars
468 .. opcode:: PK4B - Pack Four Signed 8-bit Scalars
473 .. opcode:: PK4UB - Pack Four Unsigned 8-bit Scalars
478 .. opcode:: RFL - Reflection Vector
482 dst.x = 2 \times (src0.x \times src1.x + src0.y \times src1.y + src0.z \times src1.z) / (src0.x \times src0.x + src0.y \times src0.y + src0.z \times src0.z) \times src0.x - src1.x
484 dst.y = 2 \times (src0.x \times src1.x + src0.y \times src1.y + src0.z \times src1.z) / (src0.x \times src0.x + src0.y \times src0.y + src0.z \times src0.z) \times src0.y - src1.y
486 dst.z = 2 \times (src0.x \times src1.x + src0.y \times src1.y + src0.z \times src1.z) / (src0.x \times src0.x + src0.y \times src0.y + src0.z \times src0.z) \times src0.z - src1.z
492 Considered for removal.
495 .. opcode:: SEQ - Set On Equal
499 dst.x = (src0.x == src1.x) ? 1 : 0
501 dst.y = (src0.y == src1.y) ? 1 : 0
503 dst.z = (src0.z == src1.z) ? 1 : 0
505 dst.w = (src0.w == src1.w) ? 1 : 0
508 .. opcode:: SFL - Set On False
510 This instruction replicates its result.
518 Considered for removal.
521 .. opcode:: SGT - Set On Greater Than
525 dst.x = (src0.x > src1.x) ? 1 : 0
527 dst.y = (src0.y > src1.y) ? 1 : 0
529 dst.z = (src0.z > src1.z) ? 1 : 0
531 dst.w = (src0.w > src1.w) ? 1 : 0
534 .. opcode:: SIN - Sine
536 This instruction replicates its result.
543 .. opcode:: SLE - Set On Less Equal Than
547 dst.x = (src0.x <= src1.x) ? 1 : 0
549 dst.y = (src0.y <= src1.y) ? 1 : 0
551 dst.z = (src0.z <= src1.z) ? 1 : 0
553 dst.w = (src0.w <= src1.w) ? 1 : 0
556 .. opcode:: SNE - Set On Not Equal
560 dst.x = (src0.x != src1.x) ? 1 : 0
562 dst.y = (src0.y != src1.y) ? 1 : 0
564 dst.z = (src0.z != src1.z) ? 1 : 0
566 dst.w = (src0.w != src1.w) ? 1 : 0
569 .. opcode:: STR - Set On True
571 This instruction replicates its result.
578 .. opcode:: TEX - Texture Lookup
586 dst = texture_sample(unit, coord, bias)
589 .. opcode:: TXD - Texture Lookup with Derivatives
601 dst = texture_sample_deriv(unit, coord, bias, ddx, ddy)
604 .. opcode:: TXP - Projective Texture Lookup
608 coord.x = src0.x / src.w
610 coord.y = src0.y / src.w
612 coord.z = src0.z / src.w
618 dst = texture_sample(unit, coord, bias)
621 .. opcode:: UP2H - Unpack Two 16-Bit Floats
627 Considered for removal.
629 .. opcode:: UP2US - Unpack Two Unsigned 16-Bit Scalars
635 Considered for removal.
637 .. opcode:: UP4B - Unpack Four Signed 8-Bit Values
643 Considered for removal.
645 .. opcode:: UP4UB - Unpack Four Unsigned 8-Bit Scalars
651 Considered for removal.
653 .. opcode:: X2D - 2D Coordinate Transformation
657 dst.x = src0.x + src1.x \times src2.x + src1.y \times src2.y
659 dst.y = src0.y + src1.x \times src2.z + src1.y \times src2.w
661 dst.z = src0.x + src1.x \times src2.x + src1.y \times src2.y
663 dst.w = src0.y + src1.x \times src2.z + src1.y \times src2.w
667 Considered for removal.
670 .. opcode:: ARA - Address Register Add
676 Considered for removal.
678 .. opcode:: ARR - Address Register Load With Round
691 .. opcode:: BRA - Branch
697 Considered for removal.
699 .. opcode:: CAL - Subroutine Call
705 .. opcode:: RET - Subroutine Call Return
710 .. opcode:: SSG - Set Sign
714 dst.x = (src.x > 0) ? 1 : (src.x < 0) ? -1 : 0
716 dst.y = (src.y > 0) ? 1 : (src.y < 0) ? -1 : 0
718 dst.z = (src.z > 0) ? 1 : (src.z < 0) ? -1 : 0
720 dst.w = (src.w > 0) ? 1 : (src.w < 0) ? -1 : 0
723 .. opcode:: CMP - Compare
727 dst.x = (src0.x < 0) ? src1.x : src2.x
729 dst.y = (src0.y < 0) ? src1.y : src2.y
731 dst.z = (src0.z < 0) ? src1.z : src2.z
733 dst.w = (src0.w < 0) ? src1.w : src2.w
736 .. opcode:: KIL - Conditional Discard
740 if (src.x < 0 || src.y < 0 || src.z < 0 || src.w < 0)
745 .. opcode:: SCS - Sine Cosine
758 .. opcode:: TXB - Texture Lookup With Bias
772 dst = texture_sample(unit, coord, bias)
775 .. opcode:: NRM - 3-component Vector Normalise
779 dst.x = src.x / (src.x \times src.x + src.y \times src.y + src.z \times src.z)
781 dst.y = src.y / (src.x \times src.x + src.y \times src.y + src.z \times src.z)
783 dst.z = src.z / (src.x \times src.x + src.y \times src.y + src.z \times src.z)
788 .. opcode:: DIV - Divide
792 dst.x = \frac{src0.x}{src1.x}
794 dst.y = \frac{src0.y}{src1.y}
796 dst.z = \frac{src0.z}{src1.z}
798 dst.w = \frac{src0.w}{src1.w}
801 .. opcode:: DP2 - 2-component Dot Product
803 This instruction replicates its result.
807 dst = src0.x \times src1.x + src0.y \times src1.y
810 .. opcode:: TXL - Texture Lookup With explicit LOD
824 dst = texture_sample(unit, coord, lod)
827 .. opcode:: BRK - Break
837 .. opcode:: ELSE - Else
842 .. opcode:: ENDIF - End If
847 .. opcode:: PUSHA - Push Address Register On Stack
856 Considered for cleanup.
860 Considered for removal.
862 .. opcode:: POPA - Pop Address Register From Stack
871 Considered for cleanup.
875 Considered for removal.
879 ^^^^^^^^^^^^^^^^^^^^^^^^
881 These opcodes are primarily provided for special-use computational shaders.
882 Support for these opcodes indicated by a special pipe capability bit (TBD).
884 XXX so let's discuss it, yeah?
886 .. opcode:: CEIL - Ceiling
890 dst.x = \lceil src.x\rceil
892 dst.y = \lceil src.y\rceil
894 dst.z = \lceil src.z\rceil
896 dst.w = \lceil src.w\rceil
899 .. opcode:: I2F - Integer To Float
903 dst.x = (float) src.x
905 dst.y = (float) src.y
907 dst.z = (float) src.z
909 dst.w = (float) src.w
912 .. opcode:: NOT - Bitwise Not
925 .. opcode:: TRUNC - Truncate
938 .. opcode:: SHL - Shift Left
942 dst.x = src0.x << src1.x
944 dst.y = src0.y << src1.x
946 dst.z = src0.z << src1.x
948 dst.w = src0.w << src1.x
951 .. opcode:: SHR - Shift Right
955 dst.x = src0.x >> src1.x
957 dst.y = src0.y >> src1.x
959 dst.z = src0.z >> src1.x
961 dst.w = src0.w >> src1.x
964 .. opcode:: AND - Bitwise And
968 dst.x = src0.x & src1.x
970 dst.y = src0.y & src1.y
972 dst.z = src0.z & src1.z
974 dst.w = src0.w & src1.w
977 .. opcode:: OR - Bitwise Or
981 dst.x = src0.x | src1.x
983 dst.y = src0.y | src1.y
985 dst.z = src0.z | src1.z
987 dst.w = src0.w | src1.w
990 .. opcode:: MOD - Modulus
994 dst.x = src0.x \bmod src1.x
996 dst.y = src0.y \bmod src1.y
998 dst.z = src0.z \bmod src1.z
1000 dst.w = src0.w \bmod src1.w
1003 .. opcode:: XOR - Bitwise Xor
1007 dst.x = src0.x \oplus src1.x
1009 dst.y = src0.y \oplus src1.y
1011 dst.z = src0.z \oplus src1.z
1013 dst.w = src0.w \oplus src1.w
1016 .. opcode:: SAD - Sum Of Absolute Differences
1020 dst.x = |src0.x - src1.x| + src2.x
1022 dst.y = |src0.y - src1.y| + src2.y
1024 dst.z = |src0.z - src1.z| + src2.z
1026 dst.w = |src0.w - src1.w| + src2.w
1029 .. opcode:: TXF - Texel Fetch
1034 .. opcode:: TXQ - Texture Size Query
1039 .. opcode:: CONT - Continue
1045 Support for CONT is determined by a special capability bit,
1046 ``TGSI_CONT_SUPPORTED``. See :ref:`Screen` for more information.
1050 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1052 These opcodes are only supported in geometry shaders; they have no meaning
1053 in any other type of shader.
1055 .. opcode:: EMIT - Emit
1060 .. opcode:: ENDPRIM - End Primitive
1068 These opcodes are part of :term:`GLSL`'s opcode set. Support for these
1069 opcodes is determined by a special capability bit, ``GLSL``.
1071 .. opcode:: BGNLOOP - Begin a Loop
1076 .. opcode:: BGNSUB - Begin Subroutine
1081 .. opcode:: ENDLOOP - End a Loop
1086 .. opcode:: ENDSUB - End Subroutine
1091 .. opcode:: NOP - No Operation
1096 .. opcode:: NRM4 - 4-component Vector Normalise
1098 This instruction replicates its result.
1102 dst = \frac{src.x}{src.x \times src.x + src.y \times src.y + src.z \times src.z + src.w \times src.w}
1110 .. opcode:: CALLNZ - Subroutine Call If Not Zero
1115 .. opcode:: IFC - If
1120 .. opcode:: BREAKC - Break Conditional
1129 The double-precision opcodes reinterpret four-component vectors into
1130 two-component vectors with doubled precision in each component.
1132 Support for these opcodes is XXX undecided. :T
1134 .. opcode:: DADD - Add
1138 dst.xy = src0.xy + src1.xy
1140 dst.zw = src0.zw + src1.zw
1143 .. opcode:: DDIV - Divide
1147 dst.xy = src0.xy / src1.xy
1149 dst.zw = src0.zw / src1.zw
1151 .. opcode:: DSEQ - Set on Equal
1155 dst.xy = src0.xy == src1.xy ? 1.0F : 0.0F
1157 dst.zw = src0.zw == src1.zw ? 1.0F : 0.0F
1159 .. opcode:: DSLT - Set on Less than
1163 dst.xy = src0.xy < src1.xy ? 1.0F : 0.0F
1165 dst.zw = src0.zw < src1.zw ? 1.0F : 0.0F
1167 .. opcode:: DFRAC - Fraction
1171 dst.xy = src.xy - \lfloor src.xy\rfloor
1173 dst.zw = src.zw - \lfloor src.zw\rfloor
1176 .. opcode:: DFRACEXP - Convert Number to Fractional and Integral Components
1178 Like the ``frexp()`` routine in many math libraries, this opcode stores the
1179 exponent of its source to ``dst0``, and the significand to ``dst1``, such that
1180 :math:`dst1 \times 2^{dst0} = src` .
1184 dst0.xy = exp(src.xy)
1186 dst1.xy = frac(src.xy)
1188 dst0.zw = exp(src.zw)
1190 dst1.zw = frac(src.zw)
1192 .. opcode:: DLDEXP - Multiply Number by Integral Power of 2
1194 This opcode is the inverse of :opcode:`DFRACEXP`.
1198 dst.xy = src0.xy \times 2^{src1.xy}
1200 dst.zw = src0.zw \times 2^{src1.zw}
1202 .. opcode:: DMIN - Minimum
1206 dst.xy = min(src0.xy, src1.xy)
1208 dst.zw = min(src0.zw, src1.zw)
1210 .. opcode:: DMAX - Maximum
1214 dst.xy = max(src0.xy, src1.xy)
1216 dst.zw = max(src0.zw, src1.zw)
1218 .. opcode:: DMUL - Multiply
1222 dst.xy = src0.xy \times src1.xy
1224 dst.zw = src0.zw \times src1.zw
1227 .. opcode:: DMAD - Multiply And Add
1231 dst.xy = src0.xy \times src1.xy + src2.xy
1233 dst.zw = src0.zw \times src1.zw + src2.zw
1236 .. opcode:: DRCP - Reciprocal
1240 dst.xy = \frac{1}{src.xy}
1242 dst.zw = \frac{1}{src.zw}
1244 .. opcode:: DSQRT - Square Root
1248 dst.xy = \sqrt{src.xy}
1250 dst.zw = \sqrt{src.zw}
1253 .. _resourceopcodes:
1255 Resource Access Opcodes
1256 ^^^^^^^^^^^^^^^^^^^^^^^^
1258 Those opcodes follow very closely semantics of the respective Direct3D
1259 instructions. If in doubt double check Direct3D documentation.
1261 .. opcode:: LOAD - Simplified alternative to the "SAMPLE" instruction.
1262 Using the provided integer address, LOAD fetches data
1263 from the specified buffer/texture without any filtering.
1264 The source data may come from any resource type other
1266 LOAD dst, address, resource
1268 LOAD TEMP[0], TEMP[1], RES[0]
1269 The 'address' is specified as unsigned integers. If the
1270 'address' is out of range [0...(# texels - 1)] the
1271 result of the fetch is always 0 in all components.
1272 As such the instruction doesn't honor address wrap
1273 modes, in cases where that behavior is desirable
1274 'sample' instruction should be used.
1275 address.w always provides an unsigned integer mipmap
1276 level. If the value is out of the range then the
1277 instruction always returns 0 in all components.
1278 address.yz are ignored for buffers and 1d textures.
1279 address.z is ignored for 1d texture arrays and 2d
1281 For 1D texture arrays address.y provides the array
1282 index (also as unsigned integer). If the value is
1283 out of the range of available array indices
1284 [0... (array size - 1)] then the opcode always returns
1285 0 in all components.
1286 For 2D texture arrays address.z provides the array
1287 index, otherwise it exhibits the same behavior as in
1288 the case for 1D texture arrays.
1289 The exeact semantics of the source address are presented
1291 resource type X Y Z W
1292 ------------- ------------------------
1293 PIPE_BUFFER x ignored
1294 PIPE_TEXTURE_1D x mpl
1295 PIPE_TEXTURE_2D x y mpl
1296 PIPE_TEXTURE_3D x y z mpl
1297 PIPE_TEXTURE_RECT x y mpl
1298 PIPE_TEXTURE_CUBE not allowed as source
1299 PIPE_TEXTURE_1D_ARRAY x idx mpl
1300 PIPE_TEXTURE_2D_ARRAY x y idx mpl
1302 Where 'mpl' is a mipmap level and 'idx' is the
1306 .. opcode:: LOAD_MS - Just like LOAD but allows fetch data from
1307 multi-sampled surfaces.
1309 .. opcode:: SAMPLE - Using provided address, sample data from the
1310 specified texture using the filtering mode identified
1311 by the gven sampler. The source data may come from
1312 any resource type other than buffers.
1313 SAMPLE dst, address, resource, sampler
1315 SAMPLE TEMP[0], TEMP[1], RES[0], SAMP[0]
1317 .. opcode:: SAMPLE_B - Just like the SAMPLE instruction with the
1318 exception that an additiona bias is applied to the
1319 level of detail computed as part of the instruction
1321 SAMPLE_B dst, address, resource, sampler, lod_bias
1323 SAMPLE_B TEMP[0], TEMP[1], RES[0], SAMP[0], TEMP[2].x
1325 .. opcode:: SAMPLE_C - Similar to the SAMPLE instruction but it
1326 performs a comparison filter. The operands to SAMPLE_C
1327 are identical to SAMPLE, except that tere is an additional
1328 float32 operand, reference value, which must be a register
1329 with single-component, or a scalar literal.
1330 SAMPLE_C makes the hardware use the current samplers
1331 compare_func (in pipe_sampler_state) to compare
1332 reference value against the red component value for the
1333 surce resource at each texel that the currently configured
1334 texture filter covers based on the provided coordinates.
1335 SAMPLE_C dst, address, resource.r, sampler, ref_value
1337 SAMPLE_C TEMP[0], TEMP[1], RES[0].r, SAMP[0], TEMP[2].x
1339 .. opcode:: SAMPLE_C_LZ - Same as SAMPLE_C, but LOD is 0 and derivatives
1340 are ignored. The LZ stands for level-zero.
1341 SAMPLE_C_LZ dst, address, resource.r, sampler, ref_value
1343 SAMPLE_C_LZ TEMP[0], TEMP[1], RES[0].r, SAMP[0], TEMP[2].x
1346 .. opcode:: SAMPLE_D - SAMPLE_D is identical to the SAMPLE opcode except
1347 that the derivatives for the source address in the x
1348 direction and the y direction are provided by extra
1350 SAMPLE_D dst, address, resource, sampler, der_x, der_y
1352 SAMPLE_D TEMP[0], TEMP[1], RES[0], SAMP[0], TEMP[2], TEMP[3]
1354 .. opcode:: SAMPLE_L - SAMPLE_L is identical to the SAMPLE opcode except
1355 that the LOD is provided directly as a scalar value,
1356 representing no anisotropy. Source addresses A channel
1358 SAMPLE_L dst, address, resource, sampler
1360 SAMPLE_L TEMP[0], TEMP[1], RES[0], SAMP[0]
1363 .. opcode:: GATHER4 - Gathers the four texels to be used in a bi-linear
1364 filtering operation and packs them into a single register.
1365 Only woth with 2D, 2D array, cubemaps, and cubemaps arrays.
1366 For 2D textures, only the addressing modes of the sampler and
1367 the top level of any mip pyramid are used. Set W to zero.
1368 It behaves like the SAMPLE instruction, but a filtered
1369 sample is not generated. The four samples that contribute
1370 to filtering are places into xyzw in cunter-clockwise order,
1371 starting with the (u,v) texture coordinate delta at the
1372 following locations (-, +), (+, +), (+, -), (-, -), where
1373 the magnitude of the deltas are half a texel.
1376 .. opcode:: RESINFO - query the dimensions of a given input buffer.
1377 dst receives width, height, depth or array size and
1378 number of mipmap levels. The dst can have a writemask
1379 which will specify what info is the caller interested
1381 RESINFO dst, src_mip_level, resource
1383 RESINFO TEMP[0], TEMP[1].x, RES[0]
1384 src_mip_level is an unsigned integer scalar. If it's
1385 out of range then returns 0 for width, height and
1386 depth/array size but the total number of mipmap is
1387 still returned correctly for the given resource.
1388 The returned width, height and depth values are for
1389 the mipmap level selected by the src_mip_level and
1390 are in the number of texels.
1391 For 1d texture array width is in dst.x, array size
1392 is in dst.y and dst.zw are always 0.
1394 .. opcode:: SAMPLE_POS - query the position of a given sample.
1395 dst receives float4 (x, y, 0, 0) indicated where the
1396 sample is located. If the resource is not a multi-sample
1397 resource and not a render target, the result is 0.
1399 .. opcode:: SAMPLE_INFO - dst receives number of samples in x.
1400 If the resource is not a multi-sample resource and
1401 not a render target, the result is 0.
1404 Explanation of symbols used
1405 ------------------------------
1412 :math:`|x|` Absolute value of `x`.
1414 :math:`\lceil x \rceil` Ceiling of `x`.
1416 clamp(x,y,z) Clamp x between y and z.
1417 (x < y) ? y : (x > z) ? z : x
1419 :math:`\lfloor x\rfloor` Floor of `x`.
1421 :math:`\log_2{x}` Logarithm of `x`, base 2.
1423 max(x,y) Maximum of x and y.
1426 min(x,y) Minimum of x and y.
1429 partialx(x) Derivative of x relative to fragment's X.
1431 partialy(x) Derivative of x relative to fragment's Y.
1433 pop() Pop from stack.
1435 :math:`x^y` `x` to the power `y`.
1437 push(x) Push x on stack.
1441 trunc(x) Truncate x, i.e. drop the fraction bits.
1448 discard Discard fragment.
1452 target Label of target instruction.
1463 Declares a register that is will be referenced as an operand in Instruction
1466 File field contains register file that is being declared and is one
1469 UsageMask field specifies which of the register components can be accessed
1470 and is one of TGSI_WRITEMASK.
1472 Interpolate field is only valid for fragment shader INPUT register files.
1473 It specifes the way input is being interpolated by the rasteriser and is one
1474 of TGSI_INTERPOLATE.
1476 If Dimension flag is set to 1, a Declaration Dimension token follows.
1478 If Semantic flag is set to 1, a Declaration Semantic token follows.
1480 CylindricalWrap bitfield is only valid for fragment shader INPUT register
1481 files. It specifies which register components should be subject to cylindrical
1482 wrapping when interpolating by the rasteriser. If TGSI_CYLINDRICAL_WRAP_X
1483 is set to 1, the X component should be interpolated according to cylindrical
1486 If file is TGSI_FILE_RESOURCE, a Declaration Resource token follows.
1489 Declaration Semantic
1490 ^^^^^^^^^^^^^^^^^^^^^^^^
1492 Vertex and fragment shader input and output registers may be labeled
1493 with semantic information consisting of a name and index.
1495 Follows Declaration token if Semantic bit is set.
1497 Since its purpose is to link a shader with other stages of the pipeline,
1498 it is valid to follow only those Declaration tokens that declare a register
1499 either in INPUT or OUTPUT file.
1501 SemanticName field contains the semantic name of the register being declared.
1502 There is no default value.
1504 SemanticIndex is an optional subscript that can be used to distinguish
1505 different register declarations with the same semantic name. The default value
1508 The meanings of the individual semantic names are explained in the following
1511 TGSI_SEMANTIC_POSITION
1512 """"""""""""""""""""""
1514 For vertex shaders, TGSI_SEMANTIC_POSITION indicates the vertex shader
1515 output register which contains the homogeneous vertex position in the clip
1516 space coordinate system. After clipping, the X, Y and Z components of the
1517 vertex will be divided by the W value to get normalized device coordinates.
1519 For fragment shaders, TGSI_SEMANTIC_POSITION is used to indicate that
1520 fragment shader input contains the fragment's window position. The X
1521 component starts at zero and always increases from left to right.
1522 The Y component starts at zero and always increases but Y=0 may either
1523 indicate the top of the window or the bottom depending on the fragment
1524 coordinate origin convention (see TGSI_PROPERTY_FS_COORD_ORIGIN).
1525 The Z coordinate ranges from 0 to 1 to represent depth from the front
1526 to the back of the Z buffer. The W component contains the reciprocol
1527 of the interpolated vertex position W component.
1529 Fragment shaders may also declare an output register with
1530 TGSI_SEMANTIC_POSITION. Only the Z component is writable. This allows
1531 the fragment shader to change the fragment's Z position.
1538 For vertex shader outputs or fragment shader inputs/outputs, this
1539 label indicates that the resister contains an R,G,B,A color.
1541 Several shader inputs/outputs may contain colors so the semantic index
1542 is used to distinguish them. For example, color[0] may be the diffuse
1543 color while color[1] may be the specular color.
1545 This label is needed so that the flat/smooth shading can be applied
1546 to the right interpolants during rasterization.
1550 TGSI_SEMANTIC_BCOLOR
1551 """"""""""""""""""""
1553 Back-facing colors are only used for back-facing polygons, and are only valid
1554 in vertex shader outputs. After rasterization, all polygons are front-facing
1555 and COLOR and BCOLOR end up occupying the same slots in the fragment shader,
1556 so all BCOLORs effectively become regular COLORs in the fragment shader.
1562 Vertex shader inputs and outputs and fragment shader inputs may be
1563 labeled with TGSI_SEMANTIC_FOG to indicate that the register contains
1564 a fog coordinate in the form (F, 0, 0, 1). Typically, the fragment
1565 shader will use the fog coordinate to compute a fog blend factor which
1566 is used to blend the normal fragment color with a constant fog color.
1568 Only the first component matters when writing from the vertex shader;
1569 the driver will ensure that the coordinate is in this format when used
1570 as a fragment shader input.
1576 Vertex shader input and output registers may be labeled with
1577 TGIS_SEMANTIC_PSIZE to indicate that the register contains a point size
1578 in the form (S, 0, 0, 1). The point size controls the width or diameter
1579 of points for rasterization. This label cannot be used in fragment
1582 When using this semantic, be sure to set the appropriate state in the
1583 :ref:`rasterizer` first.
1586 TGSI_SEMANTIC_GENERIC
1587 """""""""""""""""""""
1589 All vertex/fragment shader inputs/outputs not labeled with any other
1590 semantic label can be considered to be generic attributes. Typical
1591 uses of generic inputs/outputs are texcoords and user-defined values.
1594 TGSI_SEMANTIC_NORMAL
1595 """"""""""""""""""""
1597 Indicates that a vertex shader input is a normal vector. This is
1598 typically only used for legacy graphics APIs.
1604 This label applies to fragment shader inputs only and indicates that
1605 the register contains front/back-face information of the form (F, 0,
1606 0, 1). The first component will be positive when the fragment belongs
1607 to a front-facing polygon, and negative when the fragment belongs to a
1608 back-facing polygon.
1611 TGSI_SEMANTIC_EDGEFLAG
1612 """"""""""""""""""""""
1614 For vertex shaders, this sematic label indicates that an input or
1615 output is a boolean edge flag. The register layout is [F, x, x, x]
1616 where F is 0.0 or 1.0 and x = don't care. Normally, the vertex shader
1617 simply copies the edge flag input to the edgeflag output.
1619 Edge flags are used to control which lines or points are actually
1620 drawn when the polygon mode converts triangles/quads/polygons into
1623 TGSI_SEMANTIC_STENCIL
1624 """"""""""""""""""""""
1626 For fragment shaders, this semantic label indicates than an output
1627 is a writable stencil reference value. Only the Y component is writable.
1628 This allows the fragment shader to change the fragments stencilref value.
1631 Declaration Resource
1632 ^^^^^^^^^^^^^^^^^^^^^^^^
1634 Follows Declaration token if file is TGSI_FILE_RESOURCE.
1636 DCL RES[#], resource, type(s)
1638 Declares a shader input resource and assigns it to a RES[#]
1641 resource can be one of BUFFER, 1D, 2D, 3D, CUBE, 1DArray and
1644 type must be 1 or 4 entries (if specifying on a per-component
1645 level) out of UNORM, SNORM, SINT, UINT and FLOAT.
1649 ^^^^^^^^^^^^^^^^^^^^^^^^
1652 Properties are general directives that apply to the whole TGSI program.
1657 Specifies the fragment shader TGSI_SEMANTIC_POSITION coordinate origin.
1658 The default value is UPPER_LEFT.
1660 If UPPER_LEFT, the position will be (0,0) at the upper left corner and
1661 increase downward and rightward.
1662 If LOWER_LEFT, the position will be (0,0) at the lower left corner and
1663 increase upward and rightward.
1665 OpenGL defaults to LOWER_LEFT, and is configurable with the
1666 GL_ARB_fragment_coord_conventions extension.
1668 DirectX 9/10 use UPPER_LEFT.
1670 FS_COORD_PIXEL_CENTER
1671 """""""""""""""""""""
1673 Specifies the fragment shader TGSI_SEMANTIC_POSITION pixel center convention.
1674 The default value is HALF_INTEGER.
1676 If HALF_INTEGER, the fractionary part of the position will be 0.5
1677 If INTEGER, the fractionary part of the position will be 0.0
1679 Note that this does not affect the set of fragments generated by
1680 rasterization, which is instead controlled by gl_rasterization_rules in the
1683 OpenGL defaults to HALF_INTEGER, and is configurable with the
1684 GL_ARB_fragment_coord_conventions extension.
1686 DirectX 9 uses INTEGER.
1687 DirectX 10 uses HALF_INTEGER.
1689 FS_COLOR0_WRITES_ALL_CBUFS
1690 """"""""""""""""""""""""""
1691 Specifies that writes to the fragment shader color 0 are replicated to all
1692 bound cbufs. This facilitates OpenGL's fragColor output vs fragData[0] where
1693 fragData is directed to a single color buffer, but fragColor is broadcast.
1696 Texture Sampling and Texture Formats
1697 ------------------------------------
1699 This table shows how texture image components are returned as (x,y,z,w) tuples
1700 by TGSI texture instructions, such as :opcode:`TEX`, :opcode:`TXD`, and
1701 :opcode:`TXP`. For reference, OpenGL and Direct3D conventions are shown as
1704 +--------------------+--------------+--------------------+--------------+
1705 | Texture Components | Gallium | OpenGL | Direct3D 9 |
1706 +====================+==============+====================+==============+
1707 | R | (r, 0, 0, 1) | (r, 0, 0, 1) | (r, 1, 1, 1) |
1708 +--------------------+--------------+--------------------+--------------+
1709 | RG | (r, g, 0, 1) | (r, g, 0, 1) | (r, g, 1, 1) |
1710 +--------------------+--------------+--------------------+--------------+
1711 | RGB | (r, g, b, 1) | (r, g, b, 1) | (r, g, b, 1) |
1712 +--------------------+--------------+--------------------+--------------+
1713 | RGBA | (r, g, b, a) | (r, g, b, a) | (r, g, b, a) |
1714 +--------------------+--------------+--------------------+--------------+
1715 | A | (0, 0, 0, a) | (0, 0, 0, a) | (0, 0, 0, a) |
1716 +--------------------+--------------+--------------------+--------------+
1717 | L | (l, l, l, 1) | (l, l, l, 1) | (l, l, l, 1) |
1718 +--------------------+--------------+--------------------+--------------+
1719 | LA | (l, l, l, a) | (l, l, l, a) | (l, l, l, a) |
1720 +--------------------+--------------+--------------------+--------------+
1721 | I | (i, i, i, i) | (i, i, i, i) | N/A |
1722 +--------------------+--------------+--------------------+--------------+
1723 | UV | XXX TBD | (0, 0, 0, 1) | (u, v, 1, 1) |
1724 | | | [#envmap-bumpmap]_ | |
1725 +--------------------+--------------+--------------------+--------------+
1726 | Z | XXX TBD | (z, z, z, 1) | (0, z, 0, 1) |
1727 | | | [#depth-tex-mode]_ | |
1728 +--------------------+--------------+--------------------+--------------+
1729 | S | (s, s, s, s) | unknown | unknown |
1730 +--------------------+--------------+--------------------+--------------+
1732 .. [#envmap-bumpmap] http://www.opengl.org/registry/specs/ATI/envmap_bumpmap.txt
1733 .. [#depth-tex-mode] the default is (z, z, z, 1) but may also be (0, 0, 0, z)
1734 or (z, z, z, z) depending on the value of GL_DEPTH_TEXTURE_MODE.