e72b047dbd5dfd38455e6e61867ce1ed3f744ad3
[mesa.git] / src / gallium / docs / source / tgsi.rst
1 TGSI
2 ====
3
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
7 used by all drivers.
8
9 Basics
10 ------
11
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
16 one.
17
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`.
21
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.
25
26 Modifiers
27 ^^^^^^^^^^^^^^^
28
29 TGSI supports modifiers on inputs (as well as saturate and precise modifier
30 on instructions).
31
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
35 MAD instruction.
36
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
41 applying modifiers.
42
43 For inputs which have signed or unsigned type only the negate modifier is
44 supported.
45
46 Instruction Set
47 ---------------
48
49 Core ISA
50 ^^^^^^^^^^^^^^^^^^^^^^^^^
51
52 These opcodes are guaranteed to be available regardless of the driver being
53 used.
54
55 .. opcode:: ARL - Address Register Load
56
57 .. math::
58
59 dst.x = (int) \lfloor src.x\rfloor
60
61 dst.y = (int) \lfloor src.y\rfloor
62
63 dst.z = (int) \lfloor src.z\rfloor
64
65 dst.w = (int) \lfloor src.w\rfloor
66
67
68 .. opcode:: MOV - Move
69
70 .. math::
71
72 dst.x = src.x
73
74 dst.y = src.y
75
76 dst.z = src.z
77
78 dst.w = src.w
79
80
81 .. opcode:: LIT - Light Coefficients
82
83 .. math::
84
85 dst.x &= 1 \\
86 dst.y &= max(src.x, 0) \\
87 dst.z &= (src.x > 0) ? max(src.y, 0)^{clamp(src.w, -128, 128))} : 0 \\
88 dst.w &= 1
89
90
91 .. opcode:: RCP - Reciprocal
92
93 This instruction replicates its result.
94
95 .. math::
96
97 dst = \frac{1}{src.x}
98
99
100 .. opcode:: RSQ - Reciprocal Square Root
101
102 This instruction replicates its result. The results are undefined for src <= 0.
103
104 .. math::
105
106 dst = \frac{1}{\sqrt{src.x}}
107
108
109 .. opcode:: SQRT - Square Root
110
111 This instruction replicates its result. The results are undefined for src < 0.
112
113 .. math::
114
115 dst = {\sqrt{src.x}}
116
117
118 .. opcode:: EXP - Approximate Exponential Base 2
119
120 .. math::
121
122 dst.x &= 2^{\lfloor src.x\rfloor} \\
123 dst.y &= src.x - \lfloor src.x\rfloor \\
124 dst.z &= 2^{src.x} \\
125 dst.w &= 1
126
127
128 .. opcode:: LOG - Approximate Logarithm Base 2
129
130 .. math::
131
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|} \\
135 dst.w &= 1
136
137
138 .. opcode:: MUL - Multiply
139
140 .. math::
141
142 dst.x = src0.x \times src1.x
143
144 dst.y = src0.y \times src1.y
145
146 dst.z = src0.z \times src1.z
147
148 dst.w = src0.w \times src1.w
149
150
151 .. opcode:: ADD - Add
152
153 .. math::
154
155 dst.x = src0.x + src1.x
156
157 dst.y = src0.y + src1.y
158
159 dst.z = src0.z + src1.z
160
161 dst.w = src0.w + src1.w
162
163
164 .. opcode:: DP3 - 3-component Dot Product
165
166 This instruction replicates its result.
167
168 .. math::
169
170 dst = src0.x \times src1.x + src0.y \times src1.y + src0.z \times src1.z
171
172
173 .. opcode:: DP4 - 4-component Dot Product
174
175 This instruction replicates its result.
176
177 .. math::
178
179 dst = src0.x \times src1.x + src0.y \times src1.y + src0.z \times src1.z + src0.w \times src1.w
180
181
182 .. opcode:: DST - Distance Vector
183
184 .. math::
185
186 dst.x &= 1\\
187 dst.y &= src0.y \times src1.y\\
188 dst.z &= src0.z\\
189 dst.w &= src1.w
190
191
192 .. opcode:: MIN - Minimum
193
194 .. math::
195
196 dst.x = min(src0.x, src1.x)
197
198 dst.y = min(src0.y, src1.y)
199
200 dst.z = min(src0.z, src1.z)
201
202 dst.w = min(src0.w, src1.w)
203
204
205 .. opcode:: MAX - Maximum
206
207 .. math::
208
209 dst.x = max(src0.x, src1.x)
210
211 dst.y = max(src0.y, src1.y)
212
213 dst.z = max(src0.z, src1.z)
214
215 dst.w = max(src0.w, src1.w)
216
217
218 .. opcode:: SLT - Set On Less Than
219
220 .. math::
221
222 dst.x = (src0.x < src1.x) ? 1.0F : 0.0F
223
224 dst.y = (src0.y < src1.y) ? 1.0F : 0.0F
225
226 dst.z = (src0.z < src1.z) ? 1.0F : 0.0F
227
228 dst.w = (src0.w < src1.w) ? 1.0F : 0.0F
229
230
231 .. opcode:: SGE - Set On Greater Equal Than
232
233 .. math::
234
235 dst.x = (src0.x >= src1.x) ? 1.0F : 0.0F
236
237 dst.y = (src0.y >= src1.y) ? 1.0F : 0.0F
238
239 dst.z = (src0.z >= src1.z) ? 1.0F : 0.0F
240
241 dst.w = (src0.w >= src1.w) ? 1.0F : 0.0F
242
243
244 .. opcode:: MAD - Multiply And Add
245
246 Perform a * b + c. The implementation is free to decide whether there is an
247 intermediate rounding step or not.
248
249 .. math::
250
251 dst.x = src0.x \times src1.x + src2.x
252
253 dst.y = src0.y \times src1.y + src2.y
254
255 dst.z = src0.z \times src1.z + src2.z
256
257 dst.w = src0.w \times src1.w + src2.w
258
259
260 .. opcode:: LRP - Linear Interpolate
261
262 .. math::
263
264 dst.x = src0.x \times src1.x + (1 - src0.x) \times src2.x
265
266 dst.y = src0.y \times src1.y + (1 - src0.y) \times src2.y
267
268 dst.z = src0.z \times src1.z + (1 - src0.z) \times src2.z
269
270 dst.w = src0.w \times src1.w + (1 - src0.w) \times src2.w
271
272
273 .. opcode:: FMA - Fused Multiply-Add
274
275 Perform a * b + c with no intermediate rounding step.
276
277 .. math::
278
279 dst.x = src0.x \times src1.x + src2.x
280
281 dst.y = src0.y \times src1.y + src2.y
282
283 dst.z = src0.z \times src1.z + src2.z
284
285 dst.w = src0.w \times src1.w + src2.w
286
287
288 .. opcode:: FRC - Fraction
289
290 .. math::
291
292 dst.x = src.x - \lfloor src.x\rfloor
293
294 dst.y = src.y - \lfloor src.y\rfloor
295
296 dst.z = src.z - \lfloor src.z\rfloor
297
298 dst.w = src.w - \lfloor src.w\rfloor
299
300
301 .. opcode:: FLR - Floor
302
303 .. math::
304
305 dst.x = \lfloor src.x\rfloor
306
307 dst.y = \lfloor src.y\rfloor
308
309 dst.z = \lfloor src.z\rfloor
310
311 dst.w = \lfloor src.w\rfloor
312
313
314 .. opcode:: ROUND - Round
315
316 .. math::
317
318 dst.x = round(src.x)
319
320 dst.y = round(src.y)
321
322 dst.z = round(src.z)
323
324 dst.w = round(src.w)
325
326
327 .. opcode:: EX2 - Exponential Base 2
328
329 This instruction replicates its result.
330
331 .. math::
332
333 dst = 2^{src.x}
334
335
336 .. opcode:: LG2 - Logarithm Base 2
337
338 This instruction replicates its result.
339
340 .. math::
341
342 dst = \log_2{src.x}
343
344
345 .. opcode:: POW - Power
346
347 This instruction replicates its result.
348
349 .. math::
350
351 dst = src0.x^{src1.x}
352
353
354 .. opcode:: LDEXP - Multiply Number by Integral Power of 2
355
356 src1 is an integer.
357
358 .. math::
359
360 dst.x = src0.x * 2^{src1.x}
361 dst.y = src0.y * 2^{src1.y}
362 dst.z = src0.z * 2^{src1.z}
363 dst.w = src0.w * 2^{src1.w}
364
365
366 .. opcode:: COS - Cosine
367
368 This instruction replicates its result.
369
370 .. math::
371
372 dst = \cos{src.x}
373
374
375 .. opcode:: DDX, DDX_FINE - Derivative Relative To X
376
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.
380
381 .. math::
382
383 dst.x = partialx(src.x)
384
385 dst.y = partialx(src.y)
386
387 dst.z = partialx(src.z)
388
389 dst.w = partialx(src.w)
390
391
392 .. opcode:: DDY, DDY_FINE - Derivative Relative To Y
393
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.
397
398 .. math::
399
400 dst.x = partialy(src.x)
401
402 dst.y = partialy(src.y)
403
404 dst.z = partialy(src.z)
405
406 dst.w = partialy(src.w)
407
408
409 .. opcode:: PK2H - Pack Two 16-bit Floats
410
411 This instruction replicates its result.
412
413 .. math::
414
415 dst = f32\_to\_f16(src.x) | f32\_to\_f16(src.y) << 16
416
417
418 .. opcode:: PK2US - Pack Two Unsigned 16-bit Scalars
419
420 This instruction replicates its result.
421
422 .. math::
423
424 dst = f32\_to\_unorm16(src.x) | f32\_to\_unorm16(src.y) << 16
425
426
427 .. opcode:: PK4B - Pack Four Signed 8-bit Scalars
428
429 This instruction replicates its result.
430
431 .. math::
432
433 dst = f32\_to\_snorm8(src.x) |
434 (f32\_to\_snorm8(src.y) << 8) |
435 (f32\_to\_snorm8(src.z) << 16) |
436 (f32\_to\_snorm8(src.w) << 24)
437
438
439 .. opcode:: PK4UB - Pack Four Unsigned 8-bit Scalars
440
441 This instruction replicates its result.
442
443 .. math::
444
445 dst = f32\_to\_unorm8(src.x) |
446 (f32\_to\_unorm8(src.y) << 8) |
447 (f32\_to\_unorm8(src.z) << 16) |
448 (f32\_to\_unorm8(src.w) << 24)
449
450
451 .. opcode:: SEQ - Set On Equal
452
453 .. math::
454
455 dst.x = (src0.x == src1.x) ? 1.0F : 0.0F
456
457 dst.y = (src0.y == src1.y) ? 1.0F : 0.0F
458
459 dst.z = (src0.z == src1.z) ? 1.0F : 0.0F
460
461 dst.w = (src0.w == src1.w) ? 1.0F : 0.0F
462
463
464 .. opcode:: SGT - Set On Greater Than
465
466 .. math::
467
468 dst.x = (src0.x > src1.x) ? 1.0F : 0.0F
469
470 dst.y = (src0.y > src1.y) ? 1.0F : 0.0F
471
472 dst.z = (src0.z > src1.z) ? 1.0F : 0.0F
473
474 dst.w = (src0.w > src1.w) ? 1.0F : 0.0F
475
476
477 .. opcode:: SIN - Sine
478
479 This instruction replicates its result.
480
481 .. math::
482
483 dst = \sin{src.x}
484
485
486 .. opcode:: SLE - Set On Less Equal Than
487
488 .. math::
489
490 dst.x = (src0.x <= src1.x) ? 1.0F : 0.0F
491
492 dst.y = (src0.y <= src1.y) ? 1.0F : 0.0F
493
494 dst.z = (src0.z <= src1.z) ? 1.0F : 0.0F
495
496 dst.w = (src0.w <= src1.w) ? 1.0F : 0.0F
497
498
499 .. opcode:: SNE - Set On Not Equal
500
501 .. math::
502
503 dst.x = (src0.x != src1.x) ? 1.0F : 0.0F
504
505 dst.y = (src0.y != src1.y) ? 1.0F : 0.0F
506
507 dst.z = (src0.z != src1.z) ? 1.0F : 0.0F
508
509 dst.w = (src0.w != src1.w) ? 1.0F : 0.0F
510
511
512 .. opcode:: TEX - Texture Lookup
513
514 for array textures src0.y contains the slice for 1D,
515 and src0.z contain the slice for 2D.
516
517 for shadow textures with no arrays (and not cube map),
518 src0.z contains the reference value.
519
520 for shadow textures with arrays, src0.z contains
521 the reference value for 1D arrays, and src0.w contains
522 the reference value for 2D arrays and cube maps.
523
524 for cube map array shadow textures, the reference value
525 cannot be passed in src0.w, and TEX2 must be used instead.
526
527 .. math::
528
529 coord = src0
530
531 shadow_ref = src0.z or src0.w (optional)
532
533 unit = src1
534
535 dst = texture\_sample(unit, coord, shadow_ref)
536
537
538 .. opcode:: TEX2 - Texture Lookup (for shadow cube map arrays only)
539
540 this is the same as TEX, but uses another reg to encode the
541 reference value.
542
543 .. math::
544
545 coord = src0
546
547 shadow_ref = src1.x
548
549 unit = src2
550
551 dst = texture\_sample(unit, coord, shadow_ref)
552
553
554
555
556 .. opcode:: TXD - Texture Lookup with Derivatives
557
558 .. math::
559
560 coord = src0
561
562 ddx = src1
563
564 ddy = src2
565
566 unit = src3
567
568 dst = texture\_sample\_deriv(unit, coord, ddx, ddy)
569
570
571 .. opcode:: TXP - Projective Texture Lookup
572
573 .. math::
574
575 coord.x = src0.x / src0.w
576
577 coord.y = src0.y / src0.w
578
579 coord.z = src0.z / src0.w
580
581 coord.w = src0.w
582
583 unit = src1
584
585 dst = texture\_sample(unit, coord)
586
587
588 .. opcode:: UP2H - Unpack Two 16-Bit Floats
589
590 .. math::
591
592 dst.x = f16\_to\_f32(src0.x \& 0xffff)
593
594 dst.y = f16\_to\_f32(src0.x >> 16)
595
596 dst.z = f16\_to\_f32(src0.x \& 0xffff)
597
598 dst.w = f16\_to\_f32(src0.x >> 16)
599
600 .. note::
601
602 Considered for removal.
603
604 .. opcode:: UP2US - Unpack Two Unsigned 16-Bit Scalars
605
606 TBD
607
608 .. note::
609
610 Considered for removal.
611
612 .. opcode:: UP4B - Unpack Four Signed 8-Bit Values
613
614 TBD
615
616 .. note::
617
618 Considered for removal.
619
620 .. opcode:: UP4UB - Unpack Four Unsigned 8-Bit Scalars
621
622 TBD
623
624 .. note::
625
626 Considered for removal.
627
628
629 .. opcode:: ARR - Address Register Load With Round
630
631 .. math::
632
633 dst.x = (int) round(src.x)
634
635 dst.y = (int) round(src.y)
636
637 dst.z = (int) round(src.z)
638
639 dst.w = (int) round(src.w)
640
641
642 .. opcode:: SSG - Set Sign
643
644 .. math::
645
646 dst.x = (src.x > 0) ? 1 : (src.x < 0) ? -1 : 0
647
648 dst.y = (src.y > 0) ? 1 : (src.y < 0) ? -1 : 0
649
650 dst.z = (src.z > 0) ? 1 : (src.z < 0) ? -1 : 0
651
652 dst.w = (src.w > 0) ? 1 : (src.w < 0) ? -1 : 0
653
654
655 .. opcode:: CMP - Compare
656
657 .. math::
658
659 dst.x = (src0.x < 0) ? src1.x : src2.x
660
661 dst.y = (src0.y < 0) ? src1.y : src2.y
662
663 dst.z = (src0.z < 0) ? src1.z : src2.z
664
665 dst.w = (src0.w < 0) ? src1.w : src2.w
666
667
668 .. opcode:: KILL_IF - Conditional Discard
669
670 Conditional discard. Allowed in fragment shaders only.
671
672 .. math::
673
674 if (src.x < 0 || src.y < 0 || src.z < 0 || src.w < 0)
675 discard
676 endif
677
678
679 .. opcode:: KILL - Discard
680
681 Unconditional discard. Allowed in fragment shaders only.
682
683
684 .. opcode:: TXB - Texture Lookup With Bias
685
686 for cube map array textures and shadow cube maps, the bias value
687 cannot be passed in src0.w, and TXB2 must be used instead.
688
689 if the target is a shadow texture, the reference value is always
690 in src.z (this prevents shadow 3d and shadow 2d arrays from
691 using this instruction, but this is not needed).
692
693 .. math::
694
695 coord.x = src0.x
696
697 coord.y = src0.y
698
699 coord.z = src0.z
700
701 coord.w = none
702
703 bias = src0.w
704
705 unit = src1
706
707 dst = texture\_sample(unit, coord, bias)
708
709
710 .. opcode:: TXB2 - Texture Lookup With Bias (some cube maps only)
711
712 this is the same as TXB, but uses another reg to encode the
713 lod bias value for cube map arrays and shadow cube maps.
714 Presumably shadow 2d arrays and shadow 3d targets could use
715 this encoding too, but this is not legal.
716
717 shadow cube map arrays are neither possible nor required.
718
719 .. math::
720
721 coord = src0
722
723 bias = src1.x
724
725 unit = src2
726
727 dst = texture\_sample(unit, coord, bias)
728
729
730 .. opcode:: DIV - Divide
731
732 .. math::
733
734 dst.x = \frac{src0.x}{src1.x}
735
736 dst.y = \frac{src0.y}{src1.y}
737
738 dst.z = \frac{src0.z}{src1.z}
739
740 dst.w = \frac{src0.w}{src1.w}
741
742
743 .. opcode:: DP2 - 2-component Dot Product
744
745 This instruction replicates its result.
746
747 .. math::
748
749 dst = src0.x \times src1.x + src0.y \times src1.y
750
751
752 .. opcode:: TEX_LZ - Texture Lookup With LOD = 0
753
754 This is the same as TXL with LOD = 0. Like every texture opcode, it obeys
755 pipe_sampler_view::u.tex.first_level and pipe_sampler_state::min_lod.
756 There is no way to override those two in shaders.
757
758 .. math::
759
760 coord.x = src0.x
761
762 coord.y = src0.y
763
764 coord.z = src0.z
765
766 coord.w = none
767
768 lod = 0
769
770 unit = src1
771
772 dst = texture\_sample(unit, coord, lod)
773
774
775 .. opcode:: TXL - Texture Lookup With explicit LOD
776
777 for cube map array textures, the explicit lod value
778 cannot be passed in src0.w, and TXL2 must be used instead.
779
780 if the target is a shadow texture, the reference value is always
781 in src.z (this prevents shadow 3d / 2d array / cube targets from
782 using this instruction, but this is not needed).
783
784 .. math::
785
786 coord.x = src0.x
787
788 coord.y = src0.y
789
790 coord.z = src0.z
791
792 coord.w = none
793
794 lod = src0.w
795
796 unit = src1
797
798 dst = texture\_sample(unit, coord, lod)
799
800
801 .. opcode:: TXL2 - Texture Lookup With explicit LOD (for cube map arrays only)
802
803 this is the same as TXL, but uses another reg to encode the
804 explicit lod value.
805 Presumably shadow 3d / 2d array / cube targets could use
806 this encoding too, but this is not legal.
807
808 shadow cube map arrays are neither possible nor required.
809
810 .. math::
811
812 coord = src0
813
814 lod = src1.x
815
816 unit = src2
817
818 dst = texture\_sample(unit, coord, lod)
819
820
821 Compute ISA
822 ^^^^^^^^^^^^^^^^^^^^^^^^
823
824 These opcodes are primarily provided for special-use computational shaders.
825 Support for these opcodes indicated by a special pipe capability bit (TBD).
826
827 XXX doesn't look like most of the opcodes really belong here.
828
829 .. opcode:: CEIL - Ceiling
830
831 .. math::
832
833 dst.x = \lceil src.x\rceil
834
835 dst.y = \lceil src.y\rceil
836
837 dst.z = \lceil src.z\rceil
838
839 dst.w = \lceil src.w\rceil
840
841
842 .. opcode:: TRUNC - Truncate
843
844 .. math::
845
846 dst.x = trunc(src.x)
847
848 dst.y = trunc(src.y)
849
850 dst.z = trunc(src.z)
851
852 dst.w = trunc(src.w)
853
854
855 .. opcode:: MOD - Modulus
856
857 .. math::
858
859 dst.x = src0.x \bmod src1.x
860
861 dst.y = src0.y \bmod src1.y
862
863 dst.z = src0.z \bmod src1.z
864
865 dst.w = src0.w \bmod src1.w
866
867
868 .. opcode:: UARL - Integer Address Register Load
869
870 Moves the contents of the source register, assumed to be an integer, into the
871 destination register, which is assumed to be an address (ADDR) register.
872
873
874 .. opcode:: TXF - Texel Fetch
875
876 As per NV_gpu_shader4, extract a single texel from a specified texture
877 image or PIPE_BUFFER resource. The source sampler may not be a CUBE or
878 SHADOW. src 0 is a
879 four-component signed integer vector used to identify the single texel
880 accessed. 3 components + level. If the texture is multisampled, then
881 the fourth component indicates the sample, not the mipmap level.
882 Just like texture instructions, an optional
883 offset vector is provided, which is subject to various driver restrictions
884 (regarding range, source of offsets). This instruction ignores the sampler
885 state.
886
887 TXF(uint_vec coord, int_vec offset).
888
889
890 .. opcode:: TXQ - Texture Size Query
891
892 As per NV_gpu_program4, retrieve the dimensions of the texture depending on
893 the target. For 1D (width), 2D/RECT/CUBE (width, height), 3D (width, height,
894 depth), 1D array (width, layers), 2D array (width, height, layers).
895 Also return the number of accessible levels (last_level - first_level + 1)
896 in W.
897
898 For components which don't return a resource dimension, their value
899 is undefined.
900
901 .. math::
902
903 lod = src0.x
904
905 dst.x = texture\_width(unit, lod)
906
907 dst.y = texture\_height(unit, lod)
908
909 dst.z = texture\_depth(unit, lod)
910
911 dst.w = texture\_levels(unit)
912
913
914 .. opcode:: TXQS - Texture Samples Query
915
916 This retrieves the number of samples in the texture, and stores it
917 into the x component as an unsigned integer. The other components are
918 undefined. If the texture is not multisampled, this function returns
919 (1, undef, undef, undef).
920
921 .. math::
922
923 dst.x = texture\_samples(unit)
924
925
926 .. opcode:: TG4 - Texture Gather
927
928 As per ARB_texture_gather, gathers the four texels to be used in a bi-linear
929 filtering operation and packs them into a single register. Only works with
930 2D, 2D array, cubemaps, and cubemaps arrays. For 2D textures, only the
931 addressing modes of the sampler and the top level of any mip pyramid are
932 used. Set W to zero. It behaves like the TEX instruction, but a filtered
933 sample is not generated. The four samples that contribute to filtering are
934 placed into xyzw in clockwise order, starting with the (u,v) texture
935 coordinate delta at the following locations (-, +), (+, +), (+, -), (-, -),
936 where the magnitude of the deltas are half a texel.
937
938 PIPE_CAP_TEXTURE_SM5 enhances this instruction to support shadow per-sample
939 depth compares, single component selection, and a non-constant offset. It
940 doesn't allow support for the GL independent offset to get i0,j0. This would
941 require another CAP is hw can do it natively. For now we lower that before
942 TGSI.
943
944 .. math::
945
946 coord = src0
947
948 component = src1
949
950 dst = texture\_gather4 (unit, coord, component)
951
952 (with SM5 - cube array shadow)
953
954 .. math::
955
956 coord = src0
957
958 compare = src1
959
960 dst = texture\_gather (uint, coord, compare)
961
962 .. opcode:: LODQ - level of detail query
963
964 Compute the LOD information that the texture pipe would use to access the
965 texture. The Y component contains the computed LOD lambda_prime. The X
966 component contains the LOD that will be accessed, based on min/max lod's
967 and mipmap filters.
968
969 .. math::
970
971 coord = src0
972
973 dst.xy = lodq(uint, coord);
974
975 .. opcode:: CLOCK - retrieve the current shader time
976
977 Invoking this instruction multiple times in the same shader should
978 cause monotonically increasing values to be returned. The values
979 are implicitly 64-bit, so if fewer than 64 bits of precision are
980 available, to provide expected wraparound semantics, the value
981 should be shifted up so that the most significant bit of the time
982 is the most significant bit of the 64-bit value.
983
984 .. math::
985
986 dst.xy = clock()
987
988
989 Integer ISA
990 ^^^^^^^^^^^^^^^^^^^^^^^^
991 These opcodes are used for integer operations.
992 Support for these opcodes indicated by PIPE_SHADER_CAP_INTEGERS (all of them?)
993
994
995 .. opcode:: I2F - Signed Integer To Float
996
997 Rounding is unspecified (round to nearest even suggested).
998
999 .. math::
1000
1001 dst.x = (float) src.x
1002
1003 dst.y = (float) src.y
1004
1005 dst.z = (float) src.z
1006
1007 dst.w = (float) src.w
1008
1009
1010 .. opcode:: U2F - Unsigned Integer To Float
1011
1012 Rounding is unspecified (round to nearest even suggested).
1013
1014 .. math::
1015
1016 dst.x = (float) src.x
1017
1018 dst.y = (float) src.y
1019
1020 dst.z = (float) src.z
1021
1022 dst.w = (float) src.w
1023
1024
1025 .. opcode:: F2I - Float to Signed Integer
1026
1027 Rounding is towards zero (truncate).
1028 Values outside signed range (including NaNs) produce undefined results.
1029
1030 .. math::
1031
1032 dst.x = (int) src.x
1033
1034 dst.y = (int) src.y
1035
1036 dst.z = (int) src.z
1037
1038 dst.w = (int) src.w
1039
1040
1041 .. opcode:: F2U - Float to Unsigned Integer
1042
1043 Rounding is towards zero (truncate).
1044 Values outside unsigned range (including NaNs) produce undefined results.
1045
1046 .. math::
1047
1048 dst.x = (unsigned) src.x
1049
1050 dst.y = (unsigned) src.y
1051
1052 dst.z = (unsigned) src.z
1053
1054 dst.w = (unsigned) src.w
1055
1056
1057 .. opcode:: UADD - Integer Add
1058
1059 This instruction works the same for signed and unsigned integers.
1060 The low 32bit of the result is returned.
1061
1062 .. math::
1063
1064 dst.x = src0.x + src1.x
1065
1066 dst.y = src0.y + src1.y
1067
1068 dst.z = src0.z + src1.z
1069
1070 dst.w = src0.w + src1.w
1071
1072
1073 .. opcode:: UMAD - Integer Multiply And Add
1074
1075 This instruction works the same for signed and unsigned integers.
1076 The multiplication returns the low 32bit (as does the result itself).
1077
1078 .. math::
1079
1080 dst.x = src0.x \times src1.x + src2.x
1081
1082 dst.y = src0.y \times src1.y + src2.y
1083
1084 dst.z = src0.z \times src1.z + src2.z
1085
1086 dst.w = src0.w \times src1.w + src2.w
1087
1088
1089 .. opcode:: UMUL - Integer Multiply
1090
1091 This instruction works the same for signed and unsigned integers.
1092 The low 32bit of the result is returned.
1093
1094 .. math::
1095
1096 dst.x = src0.x \times src1.x
1097
1098 dst.y = src0.y \times src1.y
1099
1100 dst.z = src0.z \times src1.z
1101
1102 dst.w = src0.w \times src1.w
1103
1104
1105 .. opcode:: IMUL_HI - Signed Integer Multiply High Bits
1106
1107 The high 32bits of the multiplication of 2 signed integers are returned.
1108
1109 .. math::
1110
1111 dst.x = (src0.x \times src1.x) >> 32
1112
1113 dst.y = (src0.y \times src1.y) >> 32
1114
1115 dst.z = (src0.z \times src1.z) >> 32
1116
1117 dst.w = (src0.w \times src1.w) >> 32
1118
1119
1120 .. opcode:: UMUL_HI - Unsigned Integer Multiply High Bits
1121
1122 The high 32bits of the multiplication of 2 unsigned integers are returned.
1123
1124 .. math::
1125
1126 dst.x = (src0.x \times src1.x) >> 32
1127
1128 dst.y = (src0.y \times src1.y) >> 32
1129
1130 dst.z = (src0.z \times src1.z) >> 32
1131
1132 dst.w = (src0.w \times src1.w) >> 32
1133
1134
1135 .. opcode:: IDIV - Signed Integer Division
1136
1137 TBD: behavior for division by zero.
1138
1139 .. math::
1140
1141 dst.x = \frac{src0.x}{src1.x}
1142
1143 dst.y = \frac{src0.y}{src1.y}
1144
1145 dst.z = \frac{src0.z}{src1.z}
1146
1147 dst.w = \frac{src0.w}{src1.w}
1148
1149
1150 .. opcode:: UDIV - Unsigned Integer Division
1151
1152 For division by zero, 0xffffffff is returned.
1153
1154 .. math::
1155
1156 dst.x = \frac{src0.x}{src1.x}
1157
1158 dst.y = \frac{src0.y}{src1.y}
1159
1160 dst.z = \frac{src0.z}{src1.z}
1161
1162 dst.w = \frac{src0.w}{src1.w}
1163
1164
1165 .. opcode:: UMOD - Unsigned Integer Remainder
1166
1167 If second arg is zero, 0xffffffff is returned.
1168
1169 .. math::
1170
1171 dst.x = src0.x \bmod src1.x
1172
1173 dst.y = src0.y \bmod src1.y
1174
1175 dst.z = src0.z \bmod src1.z
1176
1177 dst.w = src0.w \bmod src1.w
1178
1179
1180 .. opcode:: NOT - Bitwise Not
1181
1182 .. math::
1183
1184 dst.x = \sim src.x
1185
1186 dst.y = \sim src.y
1187
1188 dst.z = \sim src.z
1189
1190 dst.w = \sim src.w
1191
1192
1193 .. opcode:: AND - Bitwise And
1194
1195 .. math::
1196
1197 dst.x = src0.x \& src1.x
1198
1199 dst.y = src0.y \& src1.y
1200
1201 dst.z = src0.z \& src1.z
1202
1203 dst.w = src0.w \& src1.w
1204
1205
1206 .. opcode:: OR - Bitwise Or
1207
1208 .. math::
1209
1210 dst.x = src0.x | src1.x
1211
1212 dst.y = src0.y | src1.y
1213
1214 dst.z = src0.z | src1.z
1215
1216 dst.w = src0.w | src1.w
1217
1218
1219 .. opcode:: XOR - Bitwise Xor
1220
1221 .. math::
1222
1223 dst.x = src0.x \oplus src1.x
1224
1225 dst.y = src0.y \oplus src1.y
1226
1227 dst.z = src0.z \oplus src1.z
1228
1229 dst.w = src0.w \oplus src1.w
1230
1231
1232 .. opcode:: IMAX - Maximum of Signed Integers
1233
1234 .. math::
1235
1236 dst.x = max(src0.x, src1.x)
1237
1238 dst.y = max(src0.y, src1.y)
1239
1240 dst.z = max(src0.z, src1.z)
1241
1242 dst.w = max(src0.w, src1.w)
1243
1244
1245 .. opcode:: UMAX - Maximum of Unsigned Integers
1246
1247 .. math::
1248
1249 dst.x = max(src0.x, src1.x)
1250
1251 dst.y = max(src0.y, src1.y)
1252
1253 dst.z = max(src0.z, src1.z)
1254
1255 dst.w = max(src0.w, src1.w)
1256
1257
1258 .. opcode:: IMIN - Minimum of Signed Integers
1259
1260 .. math::
1261
1262 dst.x = min(src0.x, src1.x)
1263
1264 dst.y = min(src0.y, src1.y)
1265
1266 dst.z = min(src0.z, src1.z)
1267
1268 dst.w = min(src0.w, src1.w)
1269
1270
1271 .. opcode:: UMIN - Minimum of Unsigned Integers
1272
1273 .. math::
1274
1275 dst.x = min(src0.x, src1.x)
1276
1277 dst.y = min(src0.y, src1.y)
1278
1279 dst.z = min(src0.z, src1.z)
1280
1281 dst.w = min(src0.w, src1.w)
1282
1283
1284 .. opcode:: SHL - Shift Left
1285
1286 The shift count is masked with 0x1f before the shift is applied.
1287
1288 .. math::
1289
1290 dst.x = src0.x << (0x1f \& src1.x)
1291
1292 dst.y = src0.y << (0x1f \& src1.y)
1293
1294 dst.z = src0.z << (0x1f \& src1.z)
1295
1296 dst.w = src0.w << (0x1f \& src1.w)
1297
1298
1299 .. opcode:: ISHR - Arithmetic Shift Right (of Signed Integer)
1300
1301 The shift count is masked with 0x1f before the shift is applied.
1302
1303 .. math::
1304
1305 dst.x = src0.x >> (0x1f \& src1.x)
1306
1307 dst.y = src0.y >> (0x1f \& src1.y)
1308
1309 dst.z = src0.z >> (0x1f \& src1.z)
1310
1311 dst.w = src0.w >> (0x1f \& src1.w)
1312
1313
1314 .. opcode:: USHR - Logical Shift Right
1315
1316 The shift count is masked with 0x1f before the shift is applied.
1317
1318 .. math::
1319
1320 dst.x = src0.x >> (unsigned) (0x1f \& src1.x)
1321
1322 dst.y = src0.y >> (unsigned) (0x1f \& src1.y)
1323
1324 dst.z = src0.z >> (unsigned) (0x1f \& src1.z)
1325
1326 dst.w = src0.w >> (unsigned) (0x1f \& src1.w)
1327
1328
1329 .. opcode:: UCMP - Integer Conditional Move
1330
1331 .. math::
1332
1333 dst.x = src0.x ? src1.x : src2.x
1334
1335 dst.y = src0.y ? src1.y : src2.y
1336
1337 dst.z = src0.z ? src1.z : src2.z
1338
1339 dst.w = src0.w ? src1.w : src2.w
1340
1341
1342
1343 .. opcode:: ISSG - Integer Set Sign
1344
1345 .. math::
1346
1347 dst.x = (src0.x < 0) ? -1 : (src0.x > 0) ? 1 : 0
1348
1349 dst.y = (src0.y < 0) ? -1 : (src0.y > 0) ? 1 : 0
1350
1351 dst.z = (src0.z < 0) ? -1 : (src0.z > 0) ? 1 : 0
1352
1353 dst.w = (src0.w < 0) ? -1 : (src0.w > 0) ? 1 : 0
1354
1355
1356
1357 .. opcode:: FSLT - Float Set On Less Than (ordered)
1358
1359 Same comparison as SLT but returns integer instead of 1.0/0.0 float
1360
1361 .. math::
1362
1363 dst.x = (src0.x < src1.x) ? \sim 0 : 0
1364
1365 dst.y = (src0.y < src1.y) ? \sim 0 : 0
1366
1367 dst.z = (src0.z < src1.z) ? \sim 0 : 0
1368
1369 dst.w = (src0.w < src1.w) ? \sim 0 : 0
1370
1371
1372 .. opcode:: ISLT - Signed Integer Set On Less Than
1373
1374 .. math::
1375
1376 dst.x = (src0.x < src1.x) ? \sim 0 : 0
1377
1378 dst.y = (src0.y < src1.y) ? \sim 0 : 0
1379
1380 dst.z = (src0.z < src1.z) ? \sim 0 : 0
1381
1382 dst.w = (src0.w < src1.w) ? \sim 0 : 0
1383
1384
1385 .. opcode:: USLT - Unsigned Integer Set On Less Than
1386
1387 .. math::
1388
1389 dst.x = (src0.x < src1.x) ? \sim 0 : 0
1390
1391 dst.y = (src0.y < src1.y) ? \sim 0 : 0
1392
1393 dst.z = (src0.z < src1.z) ? \sim 0 : 0
1394
1395 dst.w = (src0.w < src1.w) ? \sim 0 : 0
1396
1397
1398 .. opcode:: FSGE - Float Set On Greater Equal Than (ordered)
1399
1400 Same comparison as SGE but returns integer instead of 1.0/0.0 float
1401
1402 .. math::
1403
1404 dst.x = (src0.x >= src1.x) ? \sim 0 : 0
1405
1406 dst.y = (src0.y >= src1.y) ? \sim 0 : 0
1407
1408 dst.z = (src0.z >= src1.z) ? \sim 0 : 0
1409
1410 dst.w = (src0.w >= src1.w) ? \sim 0 : 0
1411
1412
1413 .. opcode:: ISGE - Signed Integer Set On Greater Equal Than
1414
1415 .. math::
1416
1417 dst.x = (src0.x >= src1.x) ? \sim 0 : 0
1418
1419 dst.y = (src0.y >= src1.y) ? \sim 0 : 0
1420
1421 dst.z = (src0.z >= src1.z) ? \sim 0 : 0
1422
1423 dst.w = (src0.w >= src1.w) ? \sim 0 : 0
1424
1425
1426 .. opcode:: USGE - Unsigned Integer Set On Greater Equal Than
1427
1428 .. math::
1429
1430 dst.x = (src0.x >= src1.x) ? \sim 0 : 0
1431
1432 dst.y = (src0.y >= src1.y) ? \sim 0 : 0
1433
1434 dst.z = (src0.z >= src1.z) ? \sim 0 : 0
1435
1436 dst.w = (src0.w >= src1.w) ? \sim 0 : 0
1437
1438
1439 .. opcode:: FSEQ - Float Set On Equal (ordered)
1440
1441 Same comparison as SEQ but returns integer instead of 1.0/0.0 float
1442
1443 .. math::
1444
1445 dst.x = (src0.x == src1.x) ? \sim 0 : 0
1446
1447 dst.y = (src0.y == src1.y) ? \sim 0 : 0
1448
1449 dst.z = (src0.z == src1.z) ? \sim 0 : 0
1450
1451 dst.w = (src0.w == src1.w) ? \sim 0 : 0
1452
1453
1454 .. opcode:: USEQ - Integer Set On Equal
1455
1456 .. math::
1457
1458 dst.x = (src0.x == src1.x) ? \sim 0 : 0
1459
1460 dst.y = (src0.y == src1.y) ? \sim 0 : 0
1461
1462 dst.z = (src0.z == src1.z) ? \sim 0 : 0
1463
1464 dst.w = (src0.w == src1.w) ? \sim 0 : 0
1465
1466
1467 .. opcode:: FSNE - Float Set On Not Equal (unordered)
1468
1469 Same comparison as SNE but returns integer instead of 1.0/0.0 float
1470
1471 .. math::
1472
1473 dst.x = (src0.x != src1.x) ? \sim 0 : 0
1474
1475 dst.y = (src0.y != src1.y) ? \sim 0 : 0
1476
1477 dst.z = (src0.z != src1.z) ? \sim 0 : 0
1478
1479 dst.w = (src0.w != src1.w) ? \sim 0 : 0
1480
1481
1482 .. opcode:: USNE - Integer Set On Not Equal
1483
1484 .. math::
1485
1486 dst.x = (src0.x != src1.x) ? \sim 0 : 0
1487
1488 dst.y = (src0.y != src1.y) ? \sim 0 : 0
1489
1490 dst.z = (src0.z != src1.z) ? \sim 0 : 0
1491
1492 dst.w = (src0.w != src1.w) ? \sim 0 : 0
1493
1494
1495 .. opcode:: INEG - Integer Negate
1496
1497 Two's complement.
1498
1499 .. math::
1500
1501 dst.x = -src.x
1502
1503 dst.y = -src.y
1504
1505 dst.z = -src.z
1506
1507 dst.w = -src.w
1508
1509
1510 .. opcode:: IABS - Integer Absolute Value
1511
1512 .. math::
1513
1514 dst.x = |src.x|
1515
1516 dst.y = |src.y|
1517
1518 dst.z = |src.z|
1519
1520 dst.w = |src.w|
1521
1522 Bitwise ISA
1523 ^^^^^^^^^^^
1524 These opcodes are used for bit-level manipulation of integers.
1525
1526 .. opcode:: IBFE - Signed Bitfield Extract
1527
1528 Like GLSL bitfieldExtract. Extracts a set of bits from the input, and
1529 sign-extends them if the high bit of the extracted window is set.
1530
1531 Pseudocode::
1532
1533 def ibfe(value, offset, bits):
1534 if offset < 0 or bits < 0 or offset + bits > 32:
1535 return undefined
1536 if bits == 0: return 0
1537 # Note: >> sign-extends
1538 return (value << (32 - offset - bits)) >> (32 - bits)
1539
1540 .. opcode:: UBFE - Unsigned Bitfield Extract
1541
1542 Like GLSL bitfieldExtract. Extracts a set of bits from the input, without
1543 any sign-extension.
1544
1545 Pseudocode::
1546
1547 def ubfe(value, offset, bits):
1548 if offset < 0 or bits < 0 or offset + bits > 32:
1549 return undefined
1550 if bits == 0: return 0
1551 # Note: >> does not sign-extend
1552 return (value << (32 - offset - bits)) >> (32 - bits)
1553
1554 .. opcode:: BFI - Bitfield Insert
1555
1556 Like GLSL bitfieldInsert. Replaces a bit region of 'base' with the low bits
1557 of 'insert'.
1558
1559 Pseudocode::
1560
1561 def bfi(base, insert, offset, bits):
1562 if offset < 0 or bits < 0 or offset + bits > 32:
1563 return undefined
1564 # << defined such that mask == ~0 when bits == 32, offset == 0
1565 mask = ((1 << bits) - 1) << offset
1566 return ((insert << offset) & mask) | (base & ~mask)
1567
1568 .. opcode:: BREV - Bitfield Reverse
1569
1570 See SM5 instruction BFREV. Reverses the bits of the argument.
1571
1572 .. opcode:: POPC - Population Count
1573
1574 See SM5 instruction COUNTBITS. Counts the number of set bits in the argument.
1575
1576 .. opcode:: LSB - Index of lowest set bit
1577
1578 See SM5 instruction FIRSTBIT_LO. Computes the 0-based index of the first set
1579 bit of the argument. Returns -1 if none are set.
1580
1581 .. opcode:: IMSB - Index of highest non-sign bit
1582
1583 See SM5 instruction FIRSTBIT_SHI. Computes the 0-based index of the highest
1584 non-sign bit of the argument (i.e. highest 0 bit for negative numbers,
1585 highest 1 bit for positive numbers). Returns -1 if all bits are the same
1586 (i.e. for inputs 0 and -1).
1587
1588 .. opcode:: UMSB - Index of highest set bit
1589
1590 See SM5 instruction FIRSTBIT_HI. Computes the 0-based index of the highest
1591 set bit of the argument. Returns -1 if none are set.
1592
1593 Geometry ISA
1594 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1595
1596 These opcodes are only supported in geometry shaders; they have no meaning
1597 in any other type of shader.
1598
1599 .. opcode:: EMIT - Emit
1600
1601 Generate a new vertex for the current primitive into the specified vertex
1602 stream using the values in the output registers.
1603
1604
1605 .. opcode:: ENDPRIM - End Primitive
1606
1607 Complete the current primitive in the specified vertex stream (consisting of
1608 the emitted vertices), and start a new one.
1609
1610
1611 GLSL ISA
1612 ^^^^^^^^^^
1613
1614 These opcodes are part of :term:`GLSL`'s opcode set. Support for these
1615 opcodes is determined by a special capability bit, ``GLSL``.
1616 Some require glsl version 1.30 (UIF/SWITCH/CASE/DEFAULT/ENDSWITCH).
1617
1618 .. opcode:: CAL - Subroutine Call
1619
1620 push(pc)
1621 pc = target
1622
1623
1624 .. opcode:: RET - Subroutine Call Return
1625
1626 pc = pop()
1627
1628
1629 .. opcode:: CONT - Continue
1630
1631 Unconditionally moves the point of execution to the instruction after the
1632 last bgnloop. The instruction must appear within a bgnloop/endloop.
1633
1634 .. note::
1635
1636 Support for CONT is determined by a special capability bit,
1637 ``TGSI_CONT_SUPPORTED``. See :ref:`Screen` for more information.
1638
1639
1640 .. opcode:: BGNLOOP - Begin a Loop
1641
1642 Start a loop. Must have a matching endloop.
1643
1644
1645 .. opcode:: BGNSUB - Begin Subroutine
1646
1647 Starts definition of a subroutine. Must have a matching endsub.
1648
1649
1650 .. opcode:: ENDLOOP - End a Loop
1651
1652 End a loop started with bgnloop.
1653
1654
1655 .. opcode:: ENDSUB - End Subroutine
1656
1657 Ends definition of a subroutine.
1658
1659
1660 .. opcode:: NOP - No Operation
1661
1662 Do nothing.
1663
1664
1665 .. opcode:: BRK - Break
1666
1667 Unconditionally moves the point of execution to the instruction after the
1668 next endloop or endswitch. The instruction must appear within a loop/endloop
1669 or switch/endswitch.
1670
1671
1672 .. opcode:: IF - Float If
1673
1674 Start an IF ... ELSE .. ENDIF block. Condition evaluates to true if
1675
1676 src0.x != 0.0
1677
1678 where src0.x is interpreted as a floating point register.
1679
1680
1681 .. opcode:: UIF - Bitwise If
1682
1683 Start an UIF ... ELSE .. ENDIF block. Condition evaluates to true if
1684
1685 src0.x != 0
1686
1687 where src0.x is interpreted as an integer register.
1688
1689
1690 .. opcode:: ELSE - Else
1691
1692 Starts an else block, after an IF or UIF statement.
1693
1694
1695 .. opcode:: ENDIF - End If
1696
1697 Ends an IF or UIF block.
1698
1699
1700 .. opcode:: SWITCH - Switch
1701
1702 Starts a C-style switch expression. The switch consists of one or multiple
1703 CASE statements, and at most one DEFAULT statement. Execution of a statement
1704 ends when a BRK is hit, but just like in C falling through to other cases
1705 without a break is allowed. Similarly, DEFAULT label is allowed anywhere not
1706 just as last statement, and fallthrough is allowed into/from it.
1707 CASE src arguments are evaluated at bit level against the SWITCH src argument.
1708
1709 Example::
1710
1711 SWITCH src[0].x
1712 CASE src[0].x
1713 (some instructions here)
1714 (optional BRK here)
1715 DEFAULT
1716 (some instructions here)
1717 (optional BRK here)
1718 CASE src[0].x
1719 (some instructions here)
1720 (optional BRK here)
1721 ENDSWITCH
1722
1723
1724 .. opcode:: CASE - Switch case
1725
1726 This represents a switch case label. The src arg must be an integer immediate.
1727
1728
1729 .. opcode:: DEFAULT - Switch default
1730
1731 This represents the default case in the switch, which is taken if no other
1732 case matches.
1733
1734
1735 .. opcode:: ENDSWITCH - End of switch
1736
1737 Ends a switch expression.
1738
1739
1740 Interpolation ISA
1741 ^^^^^^^^^^^^^^^^^
1742
1743 The interpolation instructions allow an input to be interpolated in a
1744 different way than its declaration. This corresponds to the GLSL 4.00
1745 interpolateAt* functions. The first argument of each of these must come from
1746 ``TGSI_FILE_INPUT``.
1747
1748 .. opcode:: INTERP_CENTROID - Interpolate at the centroid
1749
1750 Interpolates the varying specified by src0 at the centroid
1751
1752 .. opcode:: INTERP_SAMPLE - Interpolate at the specified sample
1753
1754 Interpolates the varying specified by src0 at the sample id specified by
1755 src1.x (interpreted as an integer)
1756
1757 .. opcode:: INTERP_OFFSET - Interpolate at the specified offset
1758
1759 Interpolates the varying specified by src0 at the offset src1.xy from the
1760 pixel center (interpreted as floats)
1761
1762
1763 .. _doubleopcodes:
1764
1765 Double ISA
1766 ^^^^^^^^^^^^^^^
1767
1768 The double-precision opcodes reinterpret four-component vectors into
1769 two-component vectors with doubled precision in each component.
1770
1771 .. opcode:: DABS - Absolute
1772
1773 .. math::
1774
1775 dst.xy = |src0.xy|
1776
1777 dst.zw = |src0.zw|
1778
1779 .. opcode:: DADD - Add
1780
1781 .. math::
1782
1783 dst.xy = src0.xy + src1.xy
1784
1785 dst.zw = src0.zw + src1.zw
1786
1787 .. opcode:: DSEQ - Set on Equal
1788
1789 .. math::
1790
1791 dst.x = src0.xy == src1.xy ? \sim 0 : 0
1792
1793 dst.z = src0.zw == src1.zw ? \sim 0 : 0
1794
1795 .. opcode:: DSNE - Set on Not Equal
1796
1797 .. math::
1798
1799 dst.x = src0.xy != src1.xy ? \sim 0 : 0
1800
1801 dst.z = src0.zw != src1.zw ? \sim 0 : 0
1802
1803 .. opcode:: DSLT - Set on Less than
1804
1805 .. math::
1806
1807 dst.x = src0.xy < src1.xy ? \sim 0 : 0
1808
1809 dst.z = src0.zw < src1.zw ? \sim 0 : 0
1810
1811 .. opcode:: DSGE - Set on Greater equal
1812
1813 .. math::
1814
1815 dst.x = src0.xy >= src1.xy ? \sim 0 : 0
1816
1817 dst.z = src0.zw >= src1.zw ? \sim 0 : 0
1818
1819 .. opcode:: DFRAC - Fraction
1820
1821 .. math::
1822
1823 dst.xy = src.xy - \lfloor src.xy\rfloor
1824
1825 dst.zw = src.zw - \lfloor src.zw\rfloor
1826
1827 .. opcode:: DTRUNC - Truncate
1828
1829 .. math::
1830
1831 dst.xy = trunc(src.xy)
1832
1833 dst.zw = trunc(src.zw)
1834
1835 .. opcode:: DCEIL - Ceiling
1836
1837 .. math::
1838
1839 dst.xy = \lceil src.xy\rceil
1840
1841 dst.zw = \lceil src.zw\rceil
1842
1843 .. opcode:: DFLR - Floor
1844
1845 .. math::
1846
1847 dst.xy = \lfloor src.xy\rfloor
1848
1849 dst.zw = \lfloor src.zw\rfloor
1850
1851 .. opcode:: DROUND - Fraction
1852
1853 .. math::
1854
1855 dst.xy = round(src.xy)
1856
1857 dst.zw = round(src.zw)
1858
1859 .. opcode:: DSSG - Set Sign
1860
1861 .. math::
1862
1863 dst.xy = (src.xy > 0) ? 1.0 : (src.xy < 0) ? -1.0 : 0.0
1864
1865 dst.zw = (src.zw > 0) ? 1.0 : (src.zw < 0) ? -1.0 : 0.0
1866
1867 .. opcode:: DFRACEXP - Convert Number to Fractional and Integral Components
1868
1869 Like the ``frexp()`` routine in many math libraries, this opcode stores the
1870 exponent of its source to ``dst0``, and the significand to ``dst1``, such that
1871 :math:`dst1 \times 2^{dst0} = src` . The results are replicated across
1872 channels.
1873
1874 .. math::
1875
1876 dst0.xy = dst.zw = frac(src.xy)
1877
1878 dst1 = frac(src.xy)
1879
1880
1881 .. opcode:: DLDEXP - Multiply Number by Integral Power of 2
1882
1883 This opcode is the inverse of :opcode:`DFRACEXP`. The second
1884 source is an integer.
1885
1886 .. math::
1887
1888 dst.xy = src0.xy \times 2^{src1.x}
1889
1890 dst.zw = src0.zw \times 2^{src1.z}
1891
1892 .. opcode:: DMIN - Minimum
1893
1894 .. math::
1895
1896 dst.xy = min(src0.xy, src1.xy)
1897
1898 dst.zw = min(src0.zw, src1.zw)
1899
1900 .. opcode:: DMAX - Maximum
1901
1902 .. math::
1903
1904 dst.xy = max(src0.xy, src1.xy)
1905
1906 dst.zw = max(src0.zw, src1.zw)
1907
1908 .. opcode:: DMUL - Multiply
1909
1910 .. math::
1911
1912 dst.xy = src0.xy \times src1.xy
1913
1914 dst.zw = src0.zw \times src1.zw
1915
1916
1917 .. opcode:: DMAD - Multiply And Add
1918
1919 .. math::
1920
1921 dst.xy = src0.xy \times src1.xy + src2.xy
1922
1923 dst.zw = src0.zw \times src1.zw + src2.zw
1924
1925
1926 .. opcode:: DFMA - Fused Multiply-Add
1927
1928 Perform a * b + c with no intermediate rounding step.
1929
1930 .. math::
1931
1932 dst.xy = src0.xy \times src1.xy + src2.xy
1933
1934 dst.zw = src0.zw \times src1.zw + src2.zw
1935
1936
1937 .. opcode:: DDIV - Divide
1938
1939 .. math::
1940
1941 dst.xy = \frac{src0.xy}{src1.xy}
1942
1943 dst.zw = \frac{src0.zw}{src1.zw}
1944
1945
1946 .. opcode:: DRCP - Reciprocal
1947
1948 .. math::
1949
1950 dst.xy = \frac{1}{src.xy}
1951
1952 dst.zw = \frac{1}{src.zw}
1953
1954 .. opcode:: DSQRT - Square Root
1955
1956 .. math::
1957
1958 dst.xy = \sqrt{src.xy}
1959
1960 dst.zw = \sqrt{src.zw}
1961
1962 .. opcode:: DRSQ - Reciprocal Square Root
1963
1964 .. math::
1965
1966 dst.xy = \frac{1}{\sqrt{src.xy}}
1967
1968 dst.zw = \frac{1}{\sqrt{src.zw}}
1969
1970 .. opcode:: F2D - Float to Double
1971
1972 .. math::
1973
1974 dst.xy = double(src0.x)
1975
1976 dst.zw = double(src0.y)
1977
1978 .. opcode:: D2F - Double to Float
1979
1980 .. math::
1981
1982 dst.x = float(src0.xy)
1983
1984 dst.y = float(src0.zw)
1985
1986 .. opcode:: I2D - Int to Double
1987
1988 .. math::
1989
1990 dst.xy = double(src0.x)
1991
1992 dst.zw = double(src0.y)
1993
1994 .. opcode:: D2I - Double to Int
1995
1996 .. math::
1997
1998 dst.x = int(src0.xy)
1999
2000 dst.y = int(src0.zw)
2001
2002 .. opcode:: U2D - Unsigned Int to Double
2003
2004 .. math::
2005
2006 dst.xy = double(src0.x)
2007
2008 dst.zw = double(src0.y)
2009
2010 .. opcode:: D2U - Double to Unsigned Int
2011
2012 .. math::
2013
2014 dst.x = unsigned(src0.xy)
2015
2016 dst.y = unsigned(src0.zw)
2017
2018 64-bit Integer ISA
2019 ^^^^^^^^^^^^^^^^^^
2020
2021 The 64-bit integer opcodes reinterpret four-component vectors into
2022 two-component vectors with 64-bits in each component.
2023
2024 .. opcode:: I64ABS - 64-bit Integer Absolute Value
2025
2026 .. math::
2027
2028 dst.xy = |src0.xy|
2029
2030 dst.zw = |src0.zw|
2031
2032 .. opcode:: I64NEG - 64-bit Integer Negate
2033
2034 Two's complement.
2035
2036 .. math::
2037
2038 dst.xy = -src.xy
2039
2040 dst.zw = -src.zw
2041
2042 .. opcode:: I64SSG - 64-bit Integer Set Sign
2043
2044 .. math::
2045
2046 dst.xy = (src0.xy < 0) ? -1 : (src0.xy > 0) ? 1 : 0
2047
2048 dst.zw = (src0.zw < 0) ? -1 : (src0.zw > 0) ? 1 : 0
2049
2050 .. opcode:: U64ADD - 64-bit Integer Add
2051
2052 .. math::
2053
2054 dst.xy = src0.xy + src1.xy
2055
2056 dst.zw = src0.zw + src1.zw
2057
2058 .. opcode:: U64MUL - 64-bit Integer Multiply
2059
2060 .. math::
2061
2062 dst.xy = src0.xy * src1.xy
2063
2064 dst.zw = src0.zw * src1.zw
2065
2066 .. opcode:: U64SEQ - 64-bit Integer Set on Equal
2067
2068 .. math::
2069
2070 dst.x = src0.xy == src1.xy ? \sim 0 : 0
2071
2072 dst.z = src0.zw == src1.zw ? \sim 0 : 0
2073
2074 .. opcode:: U64SNE - 64-bit Integer Set on Not Equal
2075
2076 .. math::
2077
2078 dst.x = src0.xy != src1.xy ? \sim 0 : 0
2079
2080 dst.z = src0.zw != src1.zw ? \sim 0 : 0
2081
2082 .. opcode:: U64SLT - 64-bit Unsigned Integer Set on Less Than
2083
2084 .. math::
2085
2086 dst.x = src0.xy < src1.xy ? \sim 0 : 0
2087
2088 dst.z = src0.zw < src1.zw ? \sim 0 : 0
2089
2090 .. opcode:: U64SGE - 64-bit Unsigned Integer Set on Greater Equal
2091
2092 .. math::
2093
2094 dst.x = src0.xy >= src1.xy ? \sim 0 : 0
2095
2096 dst.z = src0.zw >= src1.zw ? \sim 0 : 0
2097
2098 .. opcode:: I64SLT - 64-bit Signed Integer Set on Less Than
2099
2100 .. math::
2101
2102 dst.x = src0.xy < src1.xy ? \sim 0 : 0
2103
2104 dst.z = src0.zw < src1.zw ? \sim 0 : 0
2105
2106 .. opcode:: I64SGE - 64-bit Signed Integer Set on Greater Equal
2107
2108 .. math::
2109
2110 dst.x = src0.xy >= src1.xy ? \sim 0 : 0
2111
2112 dst.z = src0.zw >= src1.zw ? \sim 0 : 0
2113
2114 .. opcode:: I64MIN - Minimum of 64-bit Signed Integers
2115
2116 .. math::
2117
2118 dst.xy = min(src0.xy, src1.xy)
2119
2120 dst.zw = min(src0.zw, src1.zw)
2121
2122 .. opcode:: U64MIN - Minimum of 64-bit Unsigned Integers
2123
2124 .. math::
2125
2126 dst.xy = min(src0.xy, src1.xy)
2127
2128 dst.zw = min(src0.zw, src1.zw)
2129
2130 .. opcode:: I64MAX - Maximum of 64-bit Signed Integers
2131
2132 .. math::
2133
2134 dst.xy = max(src0.xy, src1.xy)
2135
2136 dst.zw = max(src0.zw, src1.zw)
2137
2138 .. opcode:: U64MAX - Maximum of 64-bit Unsigned Integers
2139
2140 .. math::
2141
2142 dst.xy = max(src0.xy, src1.xy)
2143
2144 dst.zw = max(src0.zw, src1.zw)
2145
2146 .. opcode:: U64SHL - Shift Left 64-bit Unsigned Integer
2147
2148 The shift count is masked with 0x3f before the shift is applied.
2149
2150 .. math::
2151
2152 dst.xy = src0.xy << (0x3f \& src1.x)
2153
2154 dst.zw = src0.zw << (0x3f \& src1.y)
2155
2156 .. opcode:: I64SHR - Arithmetic Shift Right (of 64-bit Signed Integer)
2157
2158 The shift count is masked with 0x3f before the shift is applied.
2159
2160 .. math::
2161
2162 dst.xy = src0.xy >> (0x3f \& src1.x)
2163
2164 dst.zw = src0.zw >> (0x3f \& src1.y)
2165
2166 .. opcode:: U64SHR - Logical Shift Right (of 64-bit Unsigned Integer)
2167
2168 The shift count is masked with 0x3f before the shift is applied.
2169
2170 .. math::
2171
2172 dst.xy = src0.xy >> (unsigned) (0x3f \& src1.x)
2173
2174 dst.zw = src0.zw >> (unsigned) (0x3f \& src1.y)
2175
2176 .. opcode:: I64DIV - 64-bit Signed Integer Division
2177
2178 .. math::
2179
2180 dst.xy = \frac{src0.xy}{src1.xy}
2181
2182 dst.zw = \frac{src0.zw}{src1.zw}
2183
2184 .. opcode:: U64DIV - 64-bit Unsigned Integer Division
2185
2186 .. math::
2187
2188 dst.xy = \frac{src0.xy}{src1.xy}
2189
2190 dst.zw = \frac{src0.zw}{src1.zw}
2191
2192 .. opcode:: U64MOD - 64-bit Unsigned Integer Remainder
2193
2194 .. math::
2195
2196 dst.xy = src0.xy \bmod src1.xy
2197
2198 dst.zw = src0.zw \bmod src1.zw
2199
2200 .. opcode:: I64MOD - 64-bit Signed Integer Remainder
2201
2202 .. math::
2203
2204 dst.xy = src0.xy \bmod src1.xy
2205
2206 dst.zw = src0.zw \bmod src1.zw
2207
2208 .. opcode:: F2U64 - Float to 64-bit Unsigned Int
2209
2210 .. math::
2211
2212 dst.xy = (uint64_t) src0.x
2213
2214 dst.zw = (uint64_t) src0.y
2215
2216 .. opcode:: F2I64 - Float to 64-bit Int
2217
2218 .. math::
2219
2220 dst.xy = (int64_t) src0.x
2221
2222 dst.zw = (int64_t) src0.y
2223
2224 .. opcode:: U2I64 - Unsigned Integer to 64-bit Integer
2225
2226 This is a zero extension.
2227
2228 .. math::
2229
2230 dst.xy = (int64_t) src0.x
2231
2232 dst.zw = (int64_t) src0.y
2233
2234 .. opcode:: I2I64 - Signed Integer to 64-bit Integer
2235
2236 This is a sign extension.
2237
2238 .. math::
2239
2240 dst.xy = (int64_t) src0.x
2241
2242 dst.zw = (int64_t) src0.y
2243
2244 .. opcode:: D2U64 - Double to 64-bit Unsigned Int
2245
2246 .. math::
2247
2248 dst.xy = (uint64_t) src0.xy
2249
2250 dst.zw = (uint64_t) src0.zw
2251
2252 .. opcode:: D2I64 - Double to 64-bit Int
2253
2254 .. math::
2255
2256 dst.xy = (int64_t) src0.xy
2257
2258 dst.zw = (int64_t) src0.zw
2259
2260 .. opcode:: U642F - 64-bit unsigned integer to float
2261
2262 .. math::
2263
2264 dst.x = (float) src0.xy
2265
2266 dst.y = (float) src0.zw
2267
2268 .. opcode:: I642F - 64-bit Int to Float
2269
2270 .. math::
2271
2272 dst.x = (float) src0.xy
2273
2274 dst.y = (float) src0.zw
2275
2276 .. opcode:: U642D - 64-bit unsigned integer to double
2277
2278 .. math::
2279
2280 dst.xy = (double) src0.xy
2281
2282 dst.zw = (double) src0.zw
2283
2284 .. opcode:: I642D - 64-bit Int to double
2285
2286 .. math::
2287
2288 dst.xy = (double) src0.xy
2289
2290 dst.zw = (double) src0.zw
2291
2292 .. _samplingopcodes:
2293
2294 Resource Sampling Opcodes
2295 ^^^^^^^^^^^^^^^^^^^^^^^^^
2296
2297 Those opcodes follow very closely semantics of the respective Direct3D
2298 instructions. If in doubt double check Direct3D documentation.
2299 Note that the swizzle on SVIEW (src1) determines texel swizzling
2300 after lookup.
2301
2302 .. opcode:: SAMPLE
2303
2304 Using provided address, sample data from the specified texture using the
2305 filtering mode identified by the given sampler. The source data may come from
2306 any resource type other than buffers.
2307
2308 Syntax: ``SAMPLE dst, address, sampler_view, sampler``
2309
2310 Example: ``SAMPLE TEMP[0], TEMP[1], SVIEW[0], SAMP[0]``
2311
2312 .. opcode:: SAMPLE_I
2313
2314 Simplified alternative to the SAMPLE instruction. Using the provided
2315 integer address, SAMPLE_I fetches data from the specified sampler view
2316 without any filtering. The source data may come from any resource type
2317 other than CUBE.
2318
2319 Syntax: ``SAMPLE_I dst, address, sampler_view``
2320
2321 Example: ``SAMPLE_I TEMP[0], TEMP[1], SVIEW[0]``
2322
2323 The 'address' is specified as unsigned integers. If the 'address' is out of
2324 range [0...(# texels - 1)] the result of the fetch is always 0 in all
2325 components. As such the instruction doesn't honor address wrap modes, in
2326 cases where that behavior is desirable 'SAMPLE' instruction should be used.
2327 address.w always provides an unsigned integer mipmap level. If the value is
2328 out of the range then the instruction always returns 0 in all components.
2329 address.yz are ignored for buffers and 1d textures. address.z is ignored
2330 for 1d texture arrays and 2d textures.
2331
2332 For 1D texture arrays address.y provides the array index (also as unsigned
2333 integer). If the value is out of the range of available array indices
2334 [0... (array size - 1)] then the opcode always returns 0 in all components.
2335 For 2D texture arrays address.z provides the array index, otherwise it
2336 exhibits the same behavior as in the case for 1D texture arrays. The exact
2337 semantics of the source address are presented in the table below:
2338
2339 +---------------------------+----+-----+-----+---------+
2340 | resource type | X | Y | Z | W |
2341 +===========================+====+=====+=====+=========+
2342 | ``PIPE_BUFFER`` | x | | | ignored |
2343 +---------------------------+----+-----+-----+---------+
2344 | ``PIPE_TEXTURE_1D`` | x | | | mpl |
2345 +---------------------------+----+-----+-----+---------+
2346 | ``PIPE_TEXTURE_2D`` | x | y | | mpl |
2347 +---------------------------+----+-----+-----+---------+
2348 | ``PIPE_TEXTURE_3D`` | x | y | z | mpl |
2349 +---------------------------+----+-----+-----+---------+
2350 | ``PIPE_TEXTURE_RECT`` | x | y | | mpl |
2351 +---------------------------+----+-----+-----+---------+
2352 | ``PIPE_TEXTURE_CUBE`` | not allowed as source |
2353 +---------------------------+----+-----+-----+---------+
2354 | ``PIPE_TEXTURE_1D_ARRAY`` | x | idx | | mpl |
2355 +---------------------------+----+-----+-----+---------+
2356 | ``PIPE_TEXTURE_2D_ARRAY`` | x | y | idx | mpl |
2357 +---------------------------+----+-----+-----+---------+
2358
2359 Where 'mpl' is a mipmap level and 'idx' is the array index.
2360
2361 .. opcode:: SAMPLE_I_MS
2362
2363 Just like SAMPLE_I but allows fetch data from multi-sampled surfaces.
2364
2365 Syntax: ``SAMPLE_I_MS dst, address, sampler_view, sample``
2366
2367 .. opcode:: SAMPLE_B
2368
2369 Just like the SAMPLE instruction with the exception that an additional bias
2370 is applied to the level of detail computed as part of the instruction
2371 execution.
2372
2373 Syntax: ``SAMPLE_B dst, address, sampler_view, sampler, lod_bias``
2374
2375 Example: ``SAMPLE_B TEMP[0], TEMP[1], SVIEW[0], SAMP[0], TEMP[2].x``
2376
2377 .. opcode:: SAMPLE_C
2378
2379 Similar to the SAMPLE instruction but it performs a comparison filter. The
2380 operands to SAMPLE_C are identical to SAMPLE, except that there is an
2381 additional float32 operand, reference value, which must be a register with
2382 single-component, or a scalar literal. SAMPLE_C makes the hardware use the
2383 current samplers compare_func (in pipe_sampler_state) to compare reference
2384 value against the red component value for the surce resource at each texel
2385 that the currently configured texture filter covers based on the provided
2386 coordinates.
2387
2388 Syntax: ``SAMPLE_C dst, address, sampler_view.r, sampler, ref_value``
2389
2390 Example: ``SAMPLE_C TEMP[0], TEMP[1], SVIEW[0].r, SAMP[0], TEMP[2].x``
2391
2392 .. opcode:: SAMPLE_C_LZ
2393
2394 Same as SAMPLE_C, but LOD is 0 and derivatives are ignored. The LZ stands
2395 for level-zero.
2396
2397 Syntax: ``SAMPLE_C_LZ dst, address, sampler_view.r, sampler, ref_value``
2398
2399 Example: ``SAMPLE_C_LZ TEMP[0], TEMP[1], SVIEW[0].r, SAMP[0], TEMP[2].x``
2400
2401
2402 .. opcode:: SAMPLE_D
2403
2404 SAMPLE_D is identical to the SAMPLE opcode except that the derivatives for
2405 the source address in the x direction and the y direction are provided by
2406 extra parameters.
2407
2408 Syntax: ``SAMPLE_D dst, address, sampler_view, sampler, der_x, der_y``
2409
2410 Example: ``SAMPLE_D TEMP[0], TEMP[1], SVIEW[0], SAMP[0], TEMP[2], TEMP[3]``
2411
2412 .. opcode:: SAMPLE_L
2413
2414 SAMPLE_L is identical to the SAMPLE opcode except that the LOD is provided
2415 directly as a scalar value, representing no anisotropy.
2416
2417 Syntax: ``SAMPLE_L dst, address, sampler_view, sampler, explicit_lod``
2418
2419 Example: ``SAMPLE_L TEMP[0], TEMP[1], SVIEW[0], SAMP[0], TEMP[2].x``
2420
2421 .. opcode:: GATHER4
2422
2423 Gathers the four texels to be used in a bi-linear filtering operation and
2424 packs them into a single register. Only works with 2D, 2D array, cubemaps,
2425 and cubemaps arrays. For 2D textures, only the addressing modes of the
2426 sampler and the top level of any mip pyramid are used. Set W to zero. It
2427 behaves like the SAMPLE instruction, but a filtered sample is not
2428 generated. The four samples that contribute to filtering are placed into
2429 xyzw in counter-clockwise order, starting with the (u,v) texture coordinate
2430 delta at the following locations (-, +), (+, +), (+, -), (-, -), where the
2431 magnitude of the deltas are half a texel.
2432
2433
2434 .. opcode:: SVIEWINFO
2435
2436 Query the dimensions of a given sampler view. dst receives width, height,
2437 depth or array size and number of mipmap levels as int4. The dst can have a
2438 writemask which will specify what info is the caller interested in.
2439
2440 Syntax: ``SVIEWINFO dst, src_mip_level, sampler_view``
2441
2442 Example: ``SVIEWINFO TEMP[0], TEMP[1].x, SVIEW[0]``
2443
2444 src_mip_level is an unsigned integer scalar. If it's out of range then
2445 returns 0 for width, height and depth/array size but the total number of
2446 mipmap is still returned correctly for the given sampler view. The returned
2447 width, height and depth values are for the mipmap level selected by the
2448 src_mip_level and are in the number of texels. For 1d texture array width
2449 is in dst.x, array size is in dst.y and dst.z is 0. The number of mipmaps is
2450 still in dst.w. In contrast to d3d10 resinfo, there's no way in the tgsi
2451 instruction encoding to specify the return type (float/rcpfloat/uint), hence
2452 always using uint. Also, unlike the SAMPLE instructions, the swizzle on src1
2453 resinfo allowing swizzling dst values is ignored (due to the interaction
2454 with rcpfloat modifier which requires some swizzle handling in the state
2455 tracker anyway).
2456
2457 .. opcode:: SAMPLE_POS
2458
2459 Query the position of a sample in the given resource or render target
2460 when per-sample fragment shading is in effect.
2461
2462 Syntax: ``SAMPLE_POS dst, source, sample_index``
2463
2464 dst receives float4 (x, y, undef, undef) indicated where the sample is
2465 located. Sample locations are in the range [0, 1] where 0.5 is the center
2466 of the fragment.
2467
2468 source is either a sampler view (to indicate a shader resource) or temp
2469 register (to indicate the render target). The source register may have
2470 an optional swizzle to apply to the returned result
2471
2472 sample_index is an integer scalar indicating which sample position is to
2473 be queried.
2474
2475 If per-sample shading is not in effect or the source resource or render
2476 target is not multisampled, the result is (0.5, 0.5, undef, undef).
2477
2478 NOTE: no driver has implemented this opcode yet (and no state tracker
2479 emits it). This information is subject to change.
2480
2481 .. opcode:: SAMPLE_INFO
2482
2483 Query the number of samples in a multisampled resource or render target.
2484
2485 Syntax: ``SAMPLE_INFO dst, source``
2486
2487 dst receives int4 (n, 0, 0, 0) where n is the number of samples in a
2488 resource or the render target.
2489
2490 source is either a sampler view (to indicate a shader resource) or temp
2491 register (to indicate the render target). The source register may have
2492 an optional swizzle to apply to the returned result
2493
2494 If per-sample shading is not in effect or the source resource or render
2495 target is not multisampled, the result is (1, 0, 0, 0).
2496
2497 NOTE: no driver has implemented this opcode yet (and no state tracker
2498 emits it). This information is subject to change.
2499
2500 .. opcode:: LOD - level of detail
2501
2502 Same syntax as the SAMPLE opcode but instead of performing an actual
2503 texture lookup/filter, return the computed LOD information that the
2504 texture pipe would use to access the texture. The Y component contains
2505 the computed LOD lambda_prime. The X component contains the LOD that will
2506 be accessed, based on min/max lod's and mipmap filters.
2507 The Z and W components are set to 0.
2508
2509 Syntax: ``LOD dst, address, sampler_view, sampler``
2510
2511
2512 .. _resourceopcodes:
2513
2514 Resource Access Opcodes
2515 ^^^^^^^^^^^^^^^^^^^^^^^
2516
2517 For these opcodes, the resource can be a BUFFER, IMAGE, or MEMORY.
2518
2519 .. opcode:: LOAD - Fetch data from a shader buffer or image
2520
2521 Syntax: ``LOAD dst, resource, address``
2522
2523 Example: ``LOAD TEMP[0], BUFFER[0], TEMP[1]``
2524
2525 Using the provided integer address, LOAD fetches data
2526 from the specified buffer or texture without any
2527 filtering.
2528
2529 The 'address' is specified as a vector of unsigned
2530 integers. If the 'address' is out of range the result
2531 is unspecified.
2532
2533 Only the first mipmap level of a resource can be read
2534 from using this instruction.
2535
2536 For 1D or 2D texture arrays, the array index is
2537 provided as an unsigned integer in address.y or
2538 address.z, respectively. address.yz are ignored for
2539 buffers and 1D textures. address.z is ignored for 1D
2540 texture arrays and 2D textures. address.w is always
2541 ignored.
2542
2543 A swizzle suffix may be added to the resource argument
2544 this will cause the resource data to be swizzled accordingly.
2545
2546 .. opcode:: STORE - Write data to a shader resource
2547
2548 Syntax: ``STORE resource, address, src``
2549
2550 Example: ``STORE BUFFER[0], TEMP[0], TEMP[1]``
2551
2552 Using the provided integer address, STORE writes data
2553 to the specified buffer or texture.
2554
2555 The 'address' is specified as a vector of unsigned
2556 integers. If the 'address' is out of range the result
2557 is unspecified.
2558
2559 Only the first mipmap level of a resource can be
2560 written to using this instruction.
2561
2562 For 1D or 2D texture arrays, the array index is
2563 provided as an unsigned integer in address.y or
2564 address.z, respectively. address.yz are ignored for
2565 buffers and 1D textures. address.z is ignored for 1D
2566 texture arrays and 2D textures. address.w is always
2567 ignored.
2568
2569 .. opcode:: RESQ - Query information about a resource
2570
2571 Syntax: ``RESQ dst, resource``
2572
2573 Example: ``RESQ TEMP[0], BUFFER[0]``
2574
2575 Returns information about the buffer or image resource. For buffer
2576 resources, the size (in bytes) is returned in the x component. For
2577 image resources, .xyz will contain the width/height/layers of the
2578 image, while .w will contain the number of samples for multi-sampled
2579 images.
2580
2581 .. opcode:: FBFETCH - Load data from framebuffer
2582
2583 Syntax: ``FBFETCH dst, output``
2584
2585 Example: ``FBFETCH TEMP[0], OUT[0]``
2586
2587 This is only valid on ``COLOR`` semantic outputs. Returns the color
2588 of the current position in the framebuffer from before this fragment
2589 shader invocation. May return the same value from multiple calls for
2590 a particular output within a single invocation. Note that result may
2591 be undefined if a fragment is drawn multiple times without a blend
2592 barrier in between.
2593
2594
2595 .. _bindlessopcodes:
2596
2597 Bindless Opcodes
2598 ^^^^^^^^^^^^^^^^
2599
2600 These opcodes are for working with bindless sampler or image handles and
2601 require PIPE_CAP_BINDLESS_TEXTURE.
2602
2603 .. opcode:: IMG2HND - Get a bindless handle for a image
2604
2605 Syntax: ``IMG2HND dst, image``
2606
2607 Example: ``IMG2HND TEMP[0], IMAGE[0]``
2608
2609 Sets 'dst' to a bindless handle for 'image'.
2610
2611 .. opcode:: SAMP2HND - Get a bindless handle for a sampler
2612
2613 Syntax: ``SAMP2HND dst, sampler``
2614
2615 Example: ``SAMP2HND TEMP[0], SAMP[0]``
2616
2617 Sets 'dst' to a bindless handle for 'sampler'.
2618
2619
2620 .. _threadsyncopcodes:
2621
2622 Inter-thread synchronization opcodes
2623 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
2624
2625 These opcodes are intended for communication between threads running
2626 within the same compute grid. For now they're only valid in compute
2627 programs.
2628
2629 .. opcode:: BARRIER - Thread group barrier
2630
2631 ``BARRIER``
2632
2633 This opcode suspends the execution of the current thread until all
2634 the remaining threads in the working group reach the same point of
2635 the program. Results are unspecified if any of the remaining
2636 threads terminates or never reaches an executed BARRIER instruction.
2637
2638 .. opcode:: MEMBAR - Memory barrier
2639
2640 ``MEMBAR type``
2641
2642 This opcode waits for the completion of all memory accesses based on
2643 the type passed in. The type is an immediate bitfield with the following
2644 meaning:
2645
2646 Bit 0: Shader storage buffers
2647 Bit 1: Atomic buffers
2648 Bit 2: Images
2649 Bit 3: Shared memory
2650 Bit 4: Thread group
2651
2652 These may be passed in in any combination. An implementation is free to not
2653 distinguish between these as it sees fit. However these map to all the
2654 possibilities made available by GLSL.
2655
2656 .. _atomopcodes:
2657
2658 Atomic opcodes
2659 ^^^^^^^^^^^^^^
2660
2661 These opcodes provide atomic variants of some common arithmetic and
2662 logical operations. In this context atomicity means that another
2663 concurrent memory access operation that affects the same memory
2664 location is guaranteed to be performed strictly before or after the
2665 entire execution of the atomic operation. The resource may be a BUFFER,
2666 IMAGE, HWATOMIC, or MEMORY. In the case of an image, the offset works
2667 the same as for ``LOAD`` and ``STORE``, specified above. For atomic
2668 counters, the offset is an immediate index to the base hw atomic
2669 counter for this operation.
2670 These atomic operations may only be used with 32-bit integer image formats.
2671
2672 .. opcode:: ATOMUADD - Atomic integer addition
2673
2674 Syntax: ``ATOMUADD dst, resource, offset, src``
2675
2676 Example: ``ATOMUADD TEMP[0], BUFFER[0], TEMP[1], TEMP[2]``
2677
2678 The following operation is performed atomically:
2679
2680 .. math::
2681
2682 dst_x = resource[offset]
2683
2684 resource[offset] = dst_x + src_x
2685
2686
2687 .. opcode:: ATOMFADD - Atomic floating point addition
2688
2689 Syntax: ``ATOMFADD dst, resource, offset, src``
2690
2691 Example: ``ATOMFADD TEMP[0], BUFFER[0], TEMP[1], TEMP[2]``
2692
2693 The following operation is performed atomically:
2694
2695 .. math::
2696
2697 dst_x = resource[offset]
2698
2699 resource[offset] = dst_x + src_x
2700
2701
2702 .. opcode:: ATOMXCHG - Atomic exchange
2703
2704 Syntax: ``ATOMXCHG dst, resource, offset, src``
2705
2706 Example: ``ATOMXCHG TEMP[0], BUFFER[0], TEMP[1], TEMP[2]``
2707
2708 The following operation is performed atomically:
2709
2710 .. math::
2711
2712 dst_x = resource[offset]
2713
2714 resource[offset] = src_x
2715
2716
2717 .. opcode:: ATOMCAS - Atomic compare-and-exchange
2718
2719 Syntax: ``ATOMCAS dst, resource, offset, cmp, src``
2720
2721 Example: ``ATOMCAS TEMP[0], BUFFER[0], TEMP[1], TEMP[2], TEMP[3]``
2722
2723 The following operation is performed atomically:
2724
2725 .. math::
2726
2727 dst_x = resource[offset]
2728
2729 resource[offset] = (dst_x == cmp_x ? src_x : dst_x)
2730
2731
2732 .. opcode:: ATOMAND - Atomic bitwise And
2733
2734 Syntax: ``ATOMAND dst, resource, offset, src``
2735
2736 Example: ``ATOMAND TEMP[0], BUFFER[0], TEMP[1], TEMP[2]``
2737
2738 The following operation is performed atomically:
2739
2740 .. math::
2741
2742 dst_x = resource[offset]
2743
2744 resource[offset] = dst_x \& src_x
2745
2746
2747 .. opcode:: ATOMOR - Atomic bitwise Or
2748
2749 Syntax: ``ATOMOR dst, resource, offset, src``
2750
2751 Example: ``ATOMOR TEMP[0], BUFFER[0], TEMP[1], TEMP[2]``
2752
2753 The following operation is performed atomically:
2754
2755 .. math::
2756
2757 dst_x = resource[offset]
2758
2759 resource[offset] = dst_x | src_x
2760
2761
2762 .. opcode:: ATOMXOR - Atomic bitwise Xor
2763
2764 Syntax: ``ATOMXOR dst, resource, offset, src``
2765
2766 Example: ``ATOMXOR TEMP[0], BUFFER[0], TEMP[1], TEMP[2]``
2767
2768 The following operation is performed atomically:
2769
2770 .. math::
2771
2772 dst_x = resource[offset]
2773
2774 resource[offset] = dst_x \oplus src_x
2775
2776
2777 .. opcode:: ATOMUMIN - Atomic unsigned minimum
2778
2779 Syntax: ``ATOMUMIN dst, resource, offset, src``
2780
2781 Example: ``ATOMUMIN TEMP[0], BUFFER[0], TEMP[1], TEMP[2]``
2782
2783 The following operation is performed atomically:
2784
2785 .. math::
2786
2787 dst_x = resource[offset]
2788
2789 resource[offset] = (dst_x < src_x ? dst_x : src_x)
2790
2791
2792 .. opcode:: ATOMUMAX - Atomic unsigned maximum
2793
2794 Syntax: ``ATOMUMAX dst, resource, offset, src``
2795
2796 Example: ``ATOMUMAX TEMP[0], BUFFER[0], TEMP[1], TEMP[2]``
2797
2798 The following operation is performed atomically:
2799
2800 .. math::
2801
2802 dst_x = resource[offset]
2803
2804 resource[offset] = (dst_x > src_x ? dst_x : src_x)
2805
2806
2807 .. opcode:: ATOMIMIN - Atomic signed minimum
2808
2809 Syntax: ``ATOMIMIN dst, resource, offset, src``
2810
2811 Example: ``ATOMIMIN TEMP[0], BUFFER[0], TEMP[1], TEMP[2]``
2812
2813 The following operation is performed atomically:
2814
2815 .. math::
2816
2817 dst_x = resource[offset]
2818
2819 resource[offset] = (dst_x < src_x ? dst_x : src_x)
2820
2821
2822 .. opcode:: ATOMIMAX - Atomic signed maximum
2823
2824 Syntax: ``ATOMIMAX dst, resource, offset, src``
2825
2826 Example: ``ATOMIMAX TEMP[0], BUFFER[0], TEMP[1], TEMP[2]``
2827
2828 The following operation is performed atomically:
2829
2830 .. math::
2831
2832 dst_x = resource[offset]
2833
2834 resource[offset] = (dst_x > src_x ? dst_x : src_x)
2835
2836
2837 .. opcode:: ATOMINC_WRAP - Atomic increment + wrap around
2838
2839 Syntax: ``ATOMINC_WRAP dst, resource, offset, src``
2840
2841 Example: ``ATOMINC_WRAP TEMP[0], BUFFER[0], TEMP[1], TEMP[2]``
2842
2843 The following operation is performed atomically:
2844
2845 .. math::
2846
2847 dst_x = resource[offset] + 1
2848
2849 resource[offset] = dst_x <= src_x ? dst_x : 0
2850
2851
2852 .. opcode:: ATOMDEC_WRAP - Atomic decrement + wrap around
2853
2854 Syntax: ``ATOMDEC_WRAP dst, resource, offset, src``
2855
2856 Example: ``ATOMDEC_WRAP TEMP[0], BUFFER[0], TEMP[1], TEMP[2]``
2857
2858 The following operation is performed atomically:
2859
2860 .. math::
2861
2862 dst_x = resource[offset]
2863
2864 resource[offset] = (dst_x > 0 && dst_x < src_x) ? dst_x - 1 : 0
2865
2866
2867 .. _interlaneopcodes:
2868
2869 Inter-lane opcodes
2870 ^^^^^^^^^^^^^^^^^^
2871
2872 These opcodes reduce the given value across the shader invocations
2873 running in the current SIMD group. Every thread in the subgroup will receive
2874 the same result. The BALLOT operations accept a single-channel argument that
2875 is treated as a boolean and produce a 64-bit value.
2876
2877 .. opcode:: VOTE_ANY - Value is set in any of the active invocations
2878
2879 Syntax: ``VOTE_ANY dst, value``
2880
2881 Example: ``VOTE_ANY TEMP[0].x, TEMP[1].x``
2882
2883
2884 .. opcode:: VOTE_ALL - Value is set in all of the active invocations
2885
2886 Syntax: ``VOTE_ALL dst, value``
2887
2888 Example: ``VOTE_ALL TEMP[0].x, TEMP[1].x``
2889
2890
2891 .. opcode:: VOTE_EQ - Value is the same in all of the active invocations
2892
2893 Syntax: ``VOTE_EQ dst, value``
2894
2895 Example: ``VOTE_EQ TEMP[0].x, TEMP[1].x``
2896
2897
2898 .. opcode:: BALLOT - Lanemask of whether the value is set in each active
2899 invocation
2900
2901 Syntax: ``BALLOT dst, value``
2902
2903 Example: ``BALLOT TEMP[0].xy, TEMP[1].x``
2904
2905 When the argument is a constant true, this produces a bitmask of active
2906 invocations. In fragment shaders, this can include helper invocations
2907 (invocations whose outputs and writes to memory are discarded, but which
2908 are used to compute derivatives).
2909
2910
2911 .. opcode:: READ_FIRST - Broadcast the value from the first active
2912 invocation to all active lanes
2913
2914 Syntax: ``READ_FIRST dst, value``
2915
2916 Example: ``READ_FIRST TEMP[0], TEMP[1]``
2917
2918
2919 .. opcode:: READ_INVOC - Retrieve the value from the given invocation
2920 (need not be uniform)
2921
2922 Syntax: ``READ_INVOC dst, value, invocation``
2923
2924 Example: ``READ_INVOC TEMP[0].xy, TEMP[1].xy, TEMP[2].x``
2925
2926 invocation.x controls the invocation number to read from for all channels.
2927 The invocation number must be the same across all active invocations in a
2928 sub-group; otherwise, the results are undefined.
2929
2930
2931 Explanation of symbols used
2932 ------------------------------
2933
2934
2935 Functions
2936 ^^^^^^^^^^^^^^
2937
2938
2939 :math:`|x|` Absolute value of `x`.
2940
2941 :math:`\lceil x \rceil` Ceiling of `x`.
2942
2943 clamp(x,y,z) Clamp x between y and z.
2944 (x < y) ? y : (x > z) ? z : x
2945
2946 :math:`\lfloor x\rfloor` Floor of `x`.
2947
2948 :math:`\log_2{x}` Logarithm of `x`, base 2.
2949
2950 max(x,y) Maximum of x and y.
2951 (x > y) ? x : y
2952
2953 min(x,y) Minimum of x and y.
2954 (x < y) ? x : y
2955
2956 partialx(x) Derivative of x relative to fragment's X.
2957
2958 partialy(x) Derivative of x relative to fragment's Y.
2959
2960 pop() Pop from stack.
2961
2962 :math:`x^y` `x` to the power `y`.
2963
2964 push(x) Push x on stack.
2965
2966 round(x) Round x.
2967
2968 trunc(x) Truncate x, i.e. drop the fraction bits.
2969
2970
2971 Keywords
2972 ^^^^^^^^^^^^^
2973
2974
2975 discard Discard fragment.
2976
2977 pc Program counter.
2978
2979 target Label of target instruction.
2980
2981
2982 Other tokens
2983 ---------------
2984
2985
2986 Declaration
2987 ^^^^^^^^^^^
2988
2989
2990 Declares a register that is will be referenced as an operand in Instruction
2991 tokens.
2992
2993 File field contains register file that is being declared and is one
2994 of TGSI_FILE.
2995
2996 UsageMask field specifies which of the register components can be accessed
2997 and is one of TGSI_WRITEMASK.
2998
2999 The Local flag specifies that a given value isn't intended for
3000 subroutine parameter passing and, as a result, the implementation
3001 isn't required to give any guarantees of it being preserved across
3002 subroutine boundaries. As it's merely a compiler hint, the
3003 implementation is free to ignore it.
3004
3005 If Dimension flag is set to 1, a Declaration Dimension token follows.
3006
3007 If Semantic flag is set to 1, a Declaration Semantic token follows.
3008
3009 If Interpolate flag is set to 1, a Declaration Interpolate token follows.
3010
3011 If file is TGSI_FILE_RESOURCE, a Declaration Resource token follows.
3012
3013 If Array flag is set to 1, a Declaration Array token follows.
3014
3015 Array Declaration
3016 ^^^^^^^^^^^^^^^^^^^^^^^^
3017
3018 Declarations can optional have an ArrayID attribute which can be referred by
3019 indirect addressing operands. An ArrayID of zero is reserved and treated as
3020 if no ArrayID is specified.
3021
3022 If an indirect addressing operand refers to a specific declaration by using
3023 an ArrayID only the registers in this declaration are guaranteed to be
3024 accessed, accessing any register outside this declaration results in undefined
3025 behavior. Note that for compatibility the effective index is zero-based and
3026 not relative to the specified declaration
3027
3028 If no ArrayID is specified with an indirect addressing operand the whole
3029 register file might be accessed by this operand. This is strongly discouraged
3030 and will prevent packing of scalar/vec2 arrays and effective alias analysis.
3031 This is only legal for TEMP and CONST register files.
3032
3033 Declaration Semantic
3034 ^^^^^^^^^^^^^^^^^^^^^^^^
3035
3036 Vertex and fragment shader input and output registers may be labeled
3037 with semantic information consisting of a name and index.
3038
3039 Follows Declaration token if Semantic bit is set.
3040
3041 Since its purpose is to link a shader with other stages of the pipeline,
3042 it is valid to follow only those Declaration tokens that declare a register
3043 either in INPUT or OUTPUT file.
3044
3045 SemanticName field contains the semantic name of the register being declared.
3046 There is no default value.
3047
3048 SemanticIndex is an optional subscript that can be used to distinguish
3049 different register declarations with the same semantic name. The default value
3050 is 0.
3051
3052 The meanings of the individual semantic names are explained in the following
3053 sections.
3054
3055 TGSI_SEMANTIC_POSITION
3056 """"""""""""""""""""""
3057
3058 For vertex shaders, TGSI_SEMANTIC_POSITION indicates the vertex shader
3059 output register which contains the homogeneous vertex position in the clip
3060 space coordinate system. After clipping, the X, Y and Z components of the
3061 vertex will be divided by the W value to get normalized device coordinates.
3062
3063 For fragment shaders, TGSI_SEMANTIC_POSITION is used to indicate that
3064 fragment shader input (or system value, depending on which one is
3065 supported by the driver) contains the fragment's window position. The X
3066 component starts at zero and always increases from left to right.
3067 The Y component starts at zero and always increases but Y=0 may either
3068 indicate the top of the window or the bottom depending on the fragment
3069 coordinate origin convention (see TGSI_PROPERTY_FS_COORD_ORIGIN).
3070 The Z coordinate ranges from 0 to 1 to represent depth from the front
3071 to the back of the Z buffer. The W component contains the interpolated
3072 reciprocal of the vertex position W component (corresponding to gl_Fragcoord,
3073 but unlike d3d10 which interpolates the same 1/w but then gives back
3074 the reciprocal of the interpolated value).
3075
3076 Fragment shaders may also declare an output register with
3077 TGSI_SEMANTIC_POSITION. Only the Z component is writable. This allows
3078 the fragment shader to change the fragment's Z position.
3079
3080
3081
3082 TGSI_SEMANTIC_COLOR
3083 """""""""""""""""""
3084
3085 For vertex shader outputs or fragment shader inputs/outputs, this
3086 label indicates that the register contains an R,G,B,A color.
3087
3088 Several shader inputs/outputs may contain colors so the semantic index
3089 is used to distinguish them. For example, color[0] may be the diffuse
3090 color while color[1] may be the specular color.
3091
3092 This label is needed so that the flat/smooth shading can be applied
3093 to the right interpolants during rasterization.
3094
3095
3096
3097 TGSI_SEMANTIC_BCOLOR
3098 """"""""""""""""""""
3099
3100 Back-facing colors are only used for back-facing polygons, and are only valid
3101 in vertex shader outputs. After rasterization, all polygons are front-facing
3102 and COLOR and BCOLOR end up occupying the same slots in the fragment shader,
3103 so all BCOLORs effectively become regular COLORs in the fragment shader.
3104
3105
3106 TGSI_SEMANTIC_FOG
3107 """""""""""""""""
3108
3109 Vertex shader inputs and outputs and fragment shader inputs may be
3110 labeled with TGSI_SEMANTIC_FOG to indicate that the register contains
3111 a fog coordinate. Typically, the fragment shader will use the fog coordinate
3112 to compute a fog blend factor which is used to blend the normal fragment color
3113 with a constant fog color. But fog coord really is just an ordinary vec4
3114 register like regular semantics.
3115
3116
3117 TGSI_SEMANTIC_PSIZE
3118 """""""""""""""""""
3119
3120 Vertex shader input and output registers may be labeled with
3121 TGIS_SEMANTIC_PSIZE to indicate that the register contains a point size
3122 in the form (S, 0, 0, 1). The point size controls the width or diameter
3123 of points for rasterization. This label cannot be used in fragment
3124 shaders.
3125
3126 When using this semantic, be sure to set the appropriate state in the
3127 :ref:`rasterizer` first.
3128
3129
3130 TGSI_SEMANTIC_TEXCOORD
3131 """"""""""""""""""""""
3132
3133 Only available if PIPE_CAP_TGSI_TEXCOORD is exposed !
3134
3135 Vertex shader outputs and fragment shader inputs may be labeled with
3136 this semantic to make them replaceable by sprite coordinates via the
3137 sprite_coord_enable state in the :ref:`rasterizer`.
3138 The semantic index permitted with this semantic is limited to <= 7.
3139
3140 If the driver does not support TEXCOORD, sprite coordinate replacement
3141 applies to inputs with the GENERIC semantic instead.
3142
3143 The intended use case for this semantic is gl_TexCoord.
3144
3145
3146 TGSI_SEMANTIC_PCOORD
3147 """"""""""""""""""""
3148
3149 Only available if PIPE_CAP_TGSI_TEXCOORD is exposed !
3150
3151 Fragment shader inputs may be labeled with TGSI_SEMANTIC_PCOORD to indicate
3152 that the register contains sprite coordinates in the form (x, y, 0, 1), if
3153 the current primitive is a point and point sprites are enabled. Otherwise,
3154 the contents of the register are undefined.
3155
3156 The intended use case for this semantic is gl_PointCoord.
3157
3158
3159 TGSI_SEMANTIC_GENERIC
3160 """""""""""""""""""""
3161
3162 All vertex/fragment shader inputs/outputs not labeled with any other
3163 semantic label can be considered to be generic attributes. Typical
3164 uses of generic inputs/outputs are texcoords and user-defined values.
3165
3166
3167 TGSI_SEMANTIC_NORMAL
3168 """"""""""""""""""""
3169
3170 Indicates that a vertex shader input is a normal vector. This is
3171 typically only used for legacy graphics APIs.
3172
3173
3174 TGSI_SEMANTIC_FACE
3175 """"""""""""""""""
3176
3177 This label applies to fragment shader inputs (or system values,
3178 depending on which one is supported by the driver) and indicates that
3179 the register contains front/back-face information.
3180
3181 If it is an input, it will be a floating-point vector in the form (F, 0, 0, 1),
3182 where F will be positive when the fragment belongs to a front-facing polygon,
3183 and negative when the fragment belongs to a back-facing polygon.
3184
3185 If it is a system value, it will be an integer vector in the form (F, 0, 0, 1),
3186 where F is 0xffffffff when the fragment belongs to a front-facing polygon and
3187 0 when the fragment belongs to a back-facing polygon.
3188
3189
3190 TGSI_SEMANTIC_EDGEFLAG
3191 """"""""""""""""""""""
3192
3193 For vertex shaders, this sematic label indicates that an input or
3194 output is a boolean edge flag. The register layout is [F, x, x, x]
3195 where F is 0.0 or 1.0 and x = don't care. Normally, the vertex shader
3196 simply copies the edge flag input to the edgeflag output.
3197
3198 Edge flags are used to control which lines or points are actually
3199 drawn when the polygon mode converts triangles/quads/polygons into
3200 points or lines.
3201
3202
3203 TGSI_SEMANTIC_STENCIL
3204 """""""""""""""""""""
3205
3206 For fragment shaders, this semantic label indicates that an output
3207 is a writable stencil reference value. Only the Y component is writable.
3208 This allows the fragment shader to change the fragments stencilref value.
3209
3210
3211 TGSI_SEMANTIC_VIEWPORT_INDEX
3212 """"""""""""""""""""""""""""
3213
3214 For geometry shaders, this semantic label indicates that an output
3215 contains the index of the viewport (and scissor) to use.
3216 This is an integer value, and only the X component is used.
3217
3218 If PIPE_CAP_TGSI_VS_LAYER_VIEWPORT or PIPE_CAP_TGSI_TES_LAYER_VIEWPORT is
3219 supported, then this semantic label can also be used in vertex or
3220 tessellation evaluation shaders, respectively. Only the value written in the
3221 last vertex processing stage is used.
3222
3223
3224 TGSI_SEMANTIC_LAYER
3225 """""""""""""""""""
3226
3227 For geometry shaders, this semantic label indicates that an output
3228 contains the layer value to use for the color and depth/stencil surfaces.
3229 This is an integer value, and only the X component is used.
3230 (Also known as rendertarget array index.)
3231
3232 If PIPE_CAP_TGSI_VS_LAYER_VIEWPORT or PIPE_CAP_TGSI_TES_LAYER_VIEWPORT is
3233 supported, then this semantic label can also be used in vertex or
3234 tessellation evaluation shaders, respectively. Only the value written in the
3235 last vertex processing stage is used.
3236
3237
3238 TGSI_SEMANTIC_CLIPDIST
3239 """"""""""""""""""""""
3240
3241 Note this covers clipping and culling distances.
3242
3243 When components of vertex elements are identified this way, these
3244 values are each assumed to be a float32 signed distance to a plane.
3245
3246 For clip distances:
3247 Primitive setup only invokes rasterization on pixels for which
3248 the interpolated plane distances are >= 0.
3249
3250 For cull distances:
3251 Primitives will be completely discarded if the plane distance
3252 for all of the vertices in the primitive are < 0.
3253 If a vertex has a cull distance of NaN, that vertex counts as "out"
3254 (as if its < 0);
3255
3256 Multiple clip/cull planes can be implemented simultaneously, by
3257 annotating multiple components of one or more vertex elements with
3258 the above specified semantic.
3259 The limits on both clip and cull distances are bound
3260 by the PIPE_MAX_CLIP_OR_CULL_DISTANCE_COUNT define which defines
3261 the maximum number of components that can be used to hold the
3262 distances and by the PIPE_MAX_CLIP_OR_CULL_DISTANCE_ELEMENT_COUNT
3263 which specifies the maximum number of registers which can be
3264 annotated with those semantics.
3265 The properties NUM_CLIPDIST_ENABLED and NUM_CULLDIST_ENABLED
3266 are used to divide up the 2 x vec4 space between clipping and culling.
3267
3268 TGSI_SEMANTIC_SAMPLEID
3269 """"""""""""""""""""""
3270
3271 For fragment shaders, this semantic label indicates that a system value
3272 contains the current sample id (i.e. gl_SampleID) as an unsigned int.
3273 Only the X component is used. If per-sample shading is not enabled,
3274 the result is (0, undef, undef, undef).
3275
3276 Note that if the fragment shader uses this system value, the fragment
3277 shader is automatically executed at per sample frequency.
3278
3279 TGSI_SEMANTIC_SAMPLEPOS
3280 """""""""""""""""""""""
3281
3282 For fragment shaders, this semantic label indicates that a system
3283 value contains the current sample's position as float4(x, y, undef, undef)
3284 in the render target (i.e. gl_SamplePosition) when per-fragment shading
3285 is in effect. Position values are in the range [0, 1] where 0.5 is
3286 the center of the fragment.
3287
3288 Note that if the fragment shader uses this system value, the fragment
3289 shader is automatically executed at per sample frequency.
3290
3291 TGSI_SEMANTIC_SAMPLEMASK
3292 """"""""""""""""""""""""
3293
3294 For fragment shaders, this semantic label can be applied to either a
3295 shader system value input or output.
3296
3297 For a system value, the sample mask indicates the set of samples covered by
3298 the current primitive. If MSAA is not enabled, the value is (1, 0, 0, 0).
3299
3300 For an output, the sample mask is used to disable further sample processing.
3301
3302 For both, the register type is uint[4] but only the X component is used
3303 (i.e. gl_SampleMask[0]). Each bit corresponds to one sample position (up
3304 to 32x MSAA is supported).
3305
3306 TGSI_SEMANTIC_INVOCATIONID
3307 """"""""""""""""""""""""""
3308
3309 For geometry shaders, this semantic label indicates that a system value
3310 contains the current invocation id (i.e. gl_InvocationID).
3311 This is an integer value, and only the X component is used.
3312
3313 TGSI_SEMANTIC_INSTANCEID
3314 """"""""""""""""""""""""
3315
3316 For vertex shaders, this semantic label indicates that a system value contains
3317 the current instance id (i.e. gl_InstanceID). It does not include the base
3318 instance. This is an integer value, and only the X component is used.
3319
3320 TGSI_SEMANTIC_VERTEXID
3321 """"""""""""""""""""""
3322
3323 For vertex shaders, this semantic label indicates that a system value contains
3324 the current vertex id (i.e. gl_VertexID). It does (unlike in d3d10) include the
3325 base vertex. This is an integer value, and only the X component is used.
3326
3327 TGSI_SEMANTIC_VERTEXID_NOBASE
3328 """""""""""""""""""""""""""""""
3329
3330 For vertex shaders, this semantic label indicates that a system value contains
3331 the current vertex id without including the base vertex (this corresponds to
3332 d3d10 vertex id, so TGSI_SEMANTIC_VERTEXID_NOBASE + TGSI_SEMANTIC_BASEVERTEX
3333 == TGSI_SEMANTIC_VERTEXID). This is an integer value, and only the X component
3334 is used.
3335
3336 TGSI_SEMANTIC_BASEVERTEX
3337 """"""""""""""""""""""""
3338
3339 For vertex shaders, this semantic label indicates that a system value contains
3340 the base vertex (i.e. gl_BaseVertex). Note that for non-indexed draw calls,
3341 this contains the first (or start) value instead.
3342 This is an integer value, and only the X component is used.
3343
3344 TGSI_SEMANTIC_PRIMID
3345 """"""""""""""""""""
3346
3347 For geometry and fragment shaders, this semantic label indicates the value
3348 contains the primitive id (i.e. gl_PrimitiveID). This is an integer value,
3349 and only the X component is used.
3350 FIXME: This right now can be either a ordinary input or a system value...
3351
3352
3353 TGSI_SEMANTIC_PATCH
3354 """""""""""""""""""
3355
3356 For tessellation evaluation/control shaders, this semantic label indicates a
3357 generic per-patch attribute. Such semantics will not implicitly be per-vertex
3358 arrays.
3359
3360 TGSI_SEMANTIC_TESSCOORD
3361 """""""""""""""""""""""
3362
3363 For tessellation evaluation shaders, this semantic label indicates the
3364 coordinates of the vertex being processed. This is available in XYZ; W is
3365 undefined.
3366
3367 TGSI_SEMANTIC_TESSOUTER
3368 """""""""""""""""""""""
3369
3370 For tessellation evaluation/control shaders, this semantic label indicates the
3371 outer tessellation levels of the patch. Isoline tessellation will only have XY
3372 defined, triangle will have XYZ and quads will have XYZW defined. This
3373 corresponds to gl_TessLevelOuter.
3374
3375 TGSI_SEMANTIC_TESSINNER
3376 """""""""""""""""""""""
3377
3378 For tessellation evaluation/control shaders, this semantic label indicates the
3379 inner tessellation levels of the patch. The X value is only defined for
3380 triangle tessellation, while quads will have XY defined. This is entirely
3381 undefined for isoline tessellation.
3382
3383 TGSI_SEMANTIC_VERTICESIN
3384 """"""""""""""""""""""""
3385
3386 For tessellation evaluation/control shaders, this semantic label indicates the
3387 number of vertices provided in the input patch. Only the X value is defined.
3388
3389 TGSI_SEMANTIC_HELPER_INVOCATION
3390 """""""""""""""""""""""""""""""
3391
3392 For fragment shaders, this semantic indicates whether the current
3393 invocation is covered or not. Helper invocations are created in order
3394 to properly compute derivatives, however it may be desirable to skip
3395 some of the logic in those cases. See ``gl_HelperInvocation`` documentation.
3396
3397 TGSI_SEMANTIC_BASEINSTANCE
3398 """"""""""""""""""""""""""
3399
3400 For vertex shaders, the base instance argument supplied for this
3401 draw. This is an integer value, and only the X component is used.
3402
3403 TGSI_SEMANTIC_DRAWID
3404 """"""""""""""""""""
3405
3406 For vertex shaders, the zero-based index of the current draw in a
3407 ``glMultiDraw*`` invocation. This is an integer value, and only the X
3408 component is used.
3409
3410
3411 TGSI_SEMANTIC_WORK_DIM
3412 """"""""""""""""""""""
3413
3414 For compute shaders started via opencl this retrieves the work_dim
3415 parameter to the clEnqueueNDRangeKernel call with which the shader
3416 was started.
3417
3418
3419 TGSI_SEMANTIC_GRID_SIZE
3420 """""""""""""""""""""""
3421
3422 For compute shaders, this semantic indicates the maximum (x, y, z) dimensions
3423 of a grid of thread blocks.
3424
3425
3426 TGSI_SEMANTIC_BLOCK_ID
3427 """"""""""""""""""""""
3428
3429 For compute shaders, this semantic indicates the (x, y, z) coordinates of the
3430 current block inside of the grid.
3431
3432
3433 TGSI_SEMANTIC_BLOCK_SIZE
3434 """"""""""""""""""""""""
3435
3436 For compute shaders, this semantic indicates the maximum (x, y, z) dimensions
3437 of a block in threads.
3438
3439
3440 TGSI_SEMANTIC_THREAD_ID
3441 """""""""""""""""""""""
3442
3443 For compute shaders, this semantic indicates the (x, y, z) coordinates of the
3444 current thread inside of the block.
3445
3446
3447 TGSI_SEMANTIC_SUBGROUP_SIZE
3448 """""""""""""""""""""""""""
3449
3450 This semantic indicates the subgroup size for the current invocation. This is
3451 an integer of at most 64, as it indicates the width of lanemasks. It does not
3452 depend on the number of invocations that are active.
3453
3454
3455 TGSI_SEMANTIC_SUBGROUP_INVOCATION
3456 """""""""""""""""""""""""""""""""
3457
3458 The index of the current invocation within its subgroup.
3459
3460
3461 TGSI_SEMANTIC_SUBGROUP_EQ_MASK
3462 """"""""""""""""""""""""""""""
3463
3464 A bit mask of ``bit index == TGSI_SEMANTIC_SUBGROUP_INVOCATION``, i.e.
3465 ``1 << subgroup_invocation`` in arbitrary precision arithmetic.
3466
3467
3468 TGSI_SEMANTIC_SUBGROUP_GE_MASK
3469 """"""""""""""""""""""""""""""
3470
3471 A bit mask of ``bit index >= TGSI_SEMANTIC_SUBGROUP_INVOCATION``, i.e.
3472 ``((1 << (subgroup_size - subgroup_invocation)) - 1) << subgroup_invocation``
3473 in arbitrary precision arithmetic.
3474
3475
3476 TGSI_SEMANTIC_SUBGROUP_GT_MASK
3477 """"""""""""""""""""""""""""""
3478
3479 A bit mask of ``bit index > TGSI_SEMANTIC_SUBGROUP_INVOCATION``, i.e.
3480 ``((1 << (subgroup_size - subgroup_invocation - 1)) - 1) << (subgroup_invocation + 1)``
3481 in arbitrary precision arithmetic.
3482
3483
3484 TGSI_SEMANTIC_SUBGROUP_LE_MASK
3485 """"""""""""""""""""""""""""""
3486
3487 A bit mask of ``bit index <= TGSI_SEMANTIC_SUBGROUP_INVOCATION``, i.e.
3488 ``(1 << (subgroup_invocation + 1)) - 1`` in arbitrary precision arithmetic.
3489
3490
3491 TGSI_SEMANTIC_SUBGROUP_LT_MASK
3492 """"""""""""""""""""""""""""""
3493
3494 A bit mask of ``bit index < TGSI_SEMANTIC_SUBGROUP_INVOCATION``, i.e.
3495 ``(1 << subgroup_invocation) - 1`` in arbitrary precision arithmetic.
3496
3497
3498 Declaration Interpolate
3499 ^^^^^^^^^^^^^^^^^^^^^^^
3500
3501 This token is only valid for fragment shader INPUT declarations.
3502
3503 The Interpolate field specifes the way input is being interpolated by
3504 the rasteriser and is one of TGSI_INTERPOLATE_*.
3505
3506 The Location field specifies the location inside the pixel that the
3507 interpolation should be done at, one of ``TGSI_INTERPOLATE_LOC_*``. Note that
3508 when per-sample shading is enabled, the implementation may choose to
3509 interpolate at the sample irrespective of the Location field.
3510
3511 The CylindricalWrap bitfield specifies which register components
3512 should be subject to cylindrical wrapping when interpolating by the
3513 rasteriser. If TGSI_CYLINDRICAL_WRAP_X is set to 1, the X component
3514 should be interpolated according to cylindrical wrapping rules.
3515
3516
3517 Declaration Sampler View
3518 ^^^^^^^^^^^^^^^^^^^^^^^^
3519
3520 Follows Declaration token if file is TGSI_FILE_SAMPLER_VIEW.
3521
3522 DCL SVIEW[#], resource, type(s)
3523
3524 Declares a shader input sampler view and assigns it to a SVIEW[#]
3525 register.
3526
3527 resource can be one of BUFFER, 1D, 2D, 3D, 1DArray and 2DArray.
3528
3529 type must be 1 or 4 entries (if specifying on a per-component
3530 level) out of UNORM, SNORM, SINT, UINT and FLOAT.
3531
3532 For TEX\* style texture sample opcodes (as opposed to SAMPLE\* opcodes
3533 which take an explicit SVIEW[#] source register), there may be optionally
3534 SVIEW[#] declarations. In this case, the SVIEW index is implied by the
3535 SAMP index, and there must be a corresponding SVIEW[#] declaration for
3536 each SAMP[#] declaration. Drivers are free to ignore this if they wish.
3537 But note in particular that some drivers need to know the sampler type
3538 (float/int/unsigned) in order to generate the correct code, so cases
3539 where integer textures are sampled, SVIEW[#] declarations should be
3540 used.
3541
3542 NOTE: It is NOT legal to mix SAMPLE\* style opcodes and TEX\* opcodes
3543 in the same shader.
3544
3545 Declaration Resource
3546 ^^^^^^^^^^^^^^^^^^^^
3547
3548 Follows Declaration token if file is TGSI_FILE_RESOURCE.
3549
3550 DCL RES[#], resource [, WR] [, RAW]
3551
3552 Declares a shader input resource and assigns it to a RES[#]
3553 register.
3554
3555 resource can be one of BUFFER, 1D, 2D, 3D, CUBE, 1DArray and
3556 2DArray.
3557
3558 If the RAW keyword is not specified, the texture data will be
3559 subject to conversion, swizzling and scaling as required to yield
3560 the specified data type from the physical data format of the bound
3561 resource.
3562
3563 If the RAW keyword is specified, no channel conversion will be
3564 performed: the values read for each of the channels (X,Y,Z,W) will
3565 correspond to consecutive words in the same order and format
3566 they're found in memory. No element-to-address conversion will be
3567 performed either: the value of the provided X coordinate will be
3568 interpreted in byte units instead of texel units. The result of
3569 accessing a misaligned address is undefined.
3570
3571 Usage of the STORE opcode is only allowed if the WR (writable) flag
3572 is set.
3573
3574 Hardware Atomic Register File
3575 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
3576
3577 Hardware atomics are declared as a 2D array with an optional array id.
3578
3579 The first member of the dimension is the buffer resource the atomic
3580 is located in.
3581 The second member is a range into the buffer resource, either for
3582 one or multiple counters. If this is an array, the declaration will have
3583 an unique array id.
3584
3585 Each counter is 4 bytes in size, and index and ranges are in counters not bytes.
3586 DCL HWATOMIC[0][0]
3587 DCL HWATOMIC[0][1]
3588
3589 This declares two atomics, one at the start of the buffer and one in the
3590 second 4 bytes.
3591
3592 DCL HWATOMIC[0][0]
3593 DCL HWATOMIC[1][0]
3594 DCL HWATOMIC[1][1..3], ARRAY(1)
3595
3596 This declares 5 atomics, one in buffer 0 at 0,
3597 one in buffer 1 at 0, and an array of 3 atomics in
3598 the buffer 1, starting at 1.
3599
3600 Properties
3601 ^^^^^^^^^^^^^^^^^^^^^^^^
3602
3603 Properties are general directives that apply to the whole TGSI program.
3604
3605 FS_COORD_ORIGIN
3606 """""""""""""""
3607
3608 Specifies the fragment shader TGSI_SEMANTIC_POSITION coordinate origin.
3609 The default value is UPPER_LEFT.
3610
3611 If UPPER_LEFT, the position will be (0,0) at the upper left corner and
3612 increase downward and rightward.
3613 If LOWER_LEFT, the position will be (0,0) at the lower left corner and
3614 increase upward and rightward.
3615
3616 OpenGL defaults to LOWER_LEFT, and is configurable with the
3617 GL_ARB_fragment_coord_conventions extension.
3618
3619 DirectX 9/10 use UPPER_LEFT.
3620
3621 FS_COORD_PIXEL_CENTER
3622 """""""""""""""""""""
3623
3624 Specifies the fragment shader TGSI_SEMANTIC_POSITION pixel center convention.
3625 The default value is HALF_INTEGER.
3626
3627 If HALF_INTEGER, the fractionary part of the position will be 0.5
3628 If INTEGER, the fractionary part of the position will be 0.0
3629
3630 Note that this does not affect the set of fragments generated by
3631 rasterization, which is instead controlled by half_pixel_center in the
3632 rasterizer.
3633
3634 OpenGL defaults to HALF_INTEGER, and is configurable with the
3635 GL_ARB_fragment_coord_conventions extension.
3636
3637 DirectX 9 uses INTEGER.
3638 DirectX 10 uses HALF_INTEGER.
3639
3640 FS_COLOR0_WRITES_ALL_CBUFS
3641 """"""""""""""""""""""""""
3642 Specifies that writes to the fragment shader color 0 are replicated to all
3643 bound cbufs. This facilitates OpenGL's fragColor output vs fragData[0] where
3644 fragData is directed to a single color buffer, but fragColor is broadcast.
3645
3646 VS_PROHIBIT_UCPS
3647 """"""""""""""""""""""""""
3648 If this property is set on the program bound to the shader stage before the
3649 fragment shader, user clip planes should have no effect (be disabled) even if
3650 that shader does not write to any clip distance outputs and the rasterizer's
3651 clip_plane_enable is non-zero.
3652 This property is only supported by drivers that also support shader clip
3653 distance outputs.
3654 This is useful for APIs that don't have UCPs and where clip distances written
3655 by a shader cannot be disabled.
3656
3657 GS_INVOCATIONS
3658 """"""""""""""
3659
3660 Specifies the number of times a geometry shader should be executed for each
3661 input primitive. Each invocation will have a different
3662 TGSI_SEMANTIC_INVOCATIONID system value set. If not specified, assumed to
3663 be 1.
3664
3665 VS_WINDOW_SPACE_POSITION
3666 """"""""""""""""""""""""""
3667 If this property is set on the vertex shader, the TGSI_SEMANTIC_POSITION output
3668 is assumed to contain window space coordinates.
3669 Division of X,Y,Z by W and the viewport transformation are disabled, and 1/W is
3670 directly taken from the 4-th component of the shader output.
3671 Naturally, clipping is not performed on window coordinates either.
3672 The effect of this property is undefined if a geometry or tessellation shader
3673 are in use.
3674
3675 TCS_VERTICES_OUT
3676 """"""""""""""""
3677
3678 The number of vertices written by the tessellation control shader. This
3679 effectively defines the patch input size of the tessellation evaluation shader
3680 as well.
3681
3682 TES_PRIM_MODE
3683 """""""""""""
3684
3685 This sets the tessellation primitive mode, one of ``PIPE_PRIM_TRIANGLES``,
3686 ``PIPE_PRIM_QUADS``, or ``PIPE_PRIM_LINES``. (Unlike in GL, there is no
3687 separate isolines settings, the regular lines is assumed to mean isolines.)
3688
3689 TES_SPACING
3690 """""""""""
3691
3692 This sets the spacing mode of the tessellation generator, one of
3693 ``PIPE_TESS_SPACING_*``.
3694
3695 TES_VERTEX_ORDER_CW
3696 """""""""""""""""""
3697
3698 This sets the vertex order to be clockwise if the value is 1, or
3699 counter-clockwise if set to 0.
3700
3701 TES_POINT_MODE
3702 """"""""""""""
3703
3704 If set to a non-zero value, this turns on point mode for the tessellator,
3705 which means that points will be generated instead of primitives.
3706
3707 NUM_CLIPDIST_ENABLED
3708 """"""""""""""""""""
3709
3710 How many clip distance scalar outputs are enabled.
3711
3712 NUM_CULLDIST_ENABLED
3713 """"""""""""""""""""
3714
3715 How many cull distance scalar outputs are enabled.
3716
3717 FS_EARLY_DEPTH_STENCIL
3718 """"""""""""""""""""""
3719
3720 Whether depth test, stencil test, and occlusion query should run before
3721 the fragment shader (regardless of fragment shader side effects). Corresponds
3722 to GLSL early_fragment_tests.
3723
3724 NEXT_SHADER
3725 """""""""""
3726
3727 Which shader stage will MOST LIKELY follow after this shader when the shader
3728 is bound. This is only a hint to the driver and doesn't have to be precise.
3729 Only set for VS and TES.
3730
3731 CS_FIXED_BLOCK_WIDTH / HEIGHT / DEPTH
3732 """""""""""""""""""""""""""""""""""""
3733
3734 Threads per block in each dimension, if known at compile time. If the block size
3735 is known all three should be at least 1. If it is unknown they should all be set
3736 to 0 or not set.
3737
3738 MUL_ZERO_WINS
3739 """""""""""""
3740
3741 The MUL TGSI operation (FP32 multiplication) will return 0 if either
3742 of the operands are equal to 0. That means that 0 * Inf = 0. This
3743 should be set the same way for an entire pipeline. Note that this
3744 applies not only to the literal MUL TGSI opcode, but all FP32
3745 multiplications implied by other operations, such as MAD, FMA, DP2,
3746 DP3, DP4, DST, LOG, LRP, and possibly others. If there is a
3747 mismatch between shaders, then it is unspecified whether this behavior
3748 will be enabled.
3749
3750 FS_POST_DEPTH_COVERAGE
3751 """"""""""""""""""""""
3752
3753 When enabled, the input for TGSI_SEMANTIC_SAMPLEMASK will exclude samples
3754 that have failed the depth/stencil tests. This is only valid when
3755 FS_EARLY_DEPTH_STENCIL is also specified.
3756
3757
3758 Texture Sampling and Texture Formats
3759 ------------------------------------
3760
3761 This table shows how texture image components are returned as (x,y,z,w) tuples
3762 by TGSI texture instructions, such as :opcode:`TEX`, :opcode:`TXD`, and
3763 :opcode:`TXP`. For reference, OpenGL and Direct3D conventions are shown as
3764 well.
3765
3766 +--------------------+--------------+--------------------+--------------+
3767 | Texture Components | Gallium | OpenGL | Direct3D 9 |
3768 +====================+==============+====================+==============+
3769 | R | (r, 0, 0, 1) | (r, 0, 0, 1) | (r, 1, 1, 1) |
3770 +--------------------+--------------+--------------------+--------------+
3771 | RG | (r, g, 0, 1) | (r, g, 0, 1) | (r, g, 1, 1) |
3772 +--------------------+--------------+--------------------+--------------+
3773 | RGB | (r, g, b, 1) | (r, g, b, 1) | (r, g, b, 1) |
3774 +--------------------+--------------+--------------------+--------------+
3775 | RGBA | (r, g, b, a) | (r, g, b, a) | (r, g, b, a) |
3776 +--------------------+--------------+--------------------+--------------+
3777 | A | (0, 0, 0, a) | (0, 0, 0, a) | (0, 0, 0, a) |
3778 +--------------------+--------------+--------------------+--------------+
3779 | L | (l, l, l, 1) | (l, l, l, 1) | (l, l, l, 1) |
3780 +--------------------+--------------+--------------------+--------------+
3781 | LA | (l, l, l, a) | (l, l, l, a) | (l, l, l, a) |
3782 +--------------------+--------------+--------------------+--------------+
3783 | I | (i, i, i, i) | (i, i, i, i) | N/A |
3784 +--------------------+--------------+--------------------+--------------+
3785 | UV | XXX TBD | (0, 0, 0, 1) | (u, v, 1, 1) |
3786 | | | [#envmap-bumpmap]_ | |
3787 +--------------------+--------------+--------------------+--------------+
3788 | Z | XXX TBD | (z, z, z, 1) | (0, z, 0, 1) |
3789 | | | [#depth-tex-mode]_ | |
3790 +--------------------+--------------+--------------------+--------------+
3791 | S | (s, s, s, s) | unknown | unknown |
3792 +--------------------+--------------+--------------------+--------------+
3793
3794 .. [#envmap-bumpmap] http://www.opengl.org/registry/specs/ATI/envmap_bumpmap.txt
3795 .. [#depth-tex-mode] the default is (z, z, z, 1) but may also be (0, 0, 0, z)
3796 or (z, z, z, z) depending on the value of GL_DEPTH_TEXTURE_MODE.