1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
|
// SPDX-FileCopyrightText: Copyright 2021 yuzu Emulator Project
// SPDX-License-Identifier: GPL-2.0-or-later
#include <string_view>
#include "shader_recompiler/backend/glsl/emit_glsl_instructions.h"
#include "shader_recompiler/backend/glsl/glsl_emit_context.h"
#include "shader_recompiler/frontend/ir/modifiers.h"
#include "shader_recompiler/frontend/ir/value.h"
#include "shader_recompiler/profile.h"
namespace Shader::Backend::GLSL {
namespace {
std::string Texture(EmitContext& ctx, const IR::TextureInstInfo& info, const IR::Value& index) {
const auto def{info.type == TextureType::Buffer ? ctx.texture_buffers.at(info.descriptor_index)
: ctx.textures.at(info.descriptor_index)};
const auto index_offset{def.count > 1 ? fmt::format("[{}]", ctx.var_alloc.Consume(index)) : ""};
return fmt::format("tex{}{}", def.binding, index_offset);
}
std::string Image(EmitContext& ctx, const IR::TextureInstInfo& info, const IR::Value& index) {
const auto def{info.type == TextureType::Buffer ? ctx.image_buffers.at(info.descriptor_index)
: ctx.images.at(info.descriptor_index)};
const auto index_offset{def.count > 1 ? fmt::format("[{}]", ctx.var_alloc.Consume(index)) : ""};
return fmt::format("img{}{}", def.binding, index_offset);
}
bool IsTextureMsaa(EmitContext& ctx, const IR::TextureInstInfo& info) {
if (info.type == TextureType::Buffer) {
return false;
}
return ctx.info.texture_descriptors.at(info.descriptor_index).is_multisample;
}
std::string CastToIntVec(std::string_view value, const IR::TextureInstInfo& info) {
switch (info.type) {
case TextureType::Color1D:
case TextureType::Buffer:
return fmt::format("int({})", value);
case TextureType::ColorArray1D:
case TextureType::Color2D:
case TextureType::ColorArray2D:
return fmt::format("ivec2({})", value);
case TextureType::Color3D:
case TextureType::ColorCube:
return fmt::format("ivec3({})", value);
case TextureType::ColorArrayCube:
return fmt::format("ivec4({})", value);
default:
throw NotImplementedException("Integer cast for TextureType {}", info.type.Value());
}
}
std::string CoordsCastToInt(std::string_view value, const IR::TextureInstInfo& info) {
switch (info.type) {
case TextureType::Color1D:
case TextureType::Buffer:
return fmt::format("int({})", value);
case TextureType::ColorArray1D:
case TextureType::Color2D:
return fmt::format("ivec2({})", value);
case TextureType::ColorArray2D:
case TextureType::Color3D:
case TextureType::ColorCube:
return fmt::format("ivec3({})", value);
case TextureType::ColorArrayCube:
return fmt::format("ivec4({})", value);
default:
throw NotImplementedException("TexelFetchCast type {}", info.type.Value());
}
}
bool NeedsShadowLodExt(TextureType type) {
switch (type) {
case TextureType::ColorArray2D:
case TextureType::ColorCube:
case TextureType::ColorArrayCube:
return true;
default:
return false;
}
}
std::string GetOffsetVec(EmitContext& ctx, const IR::Value& offset) {
if (offset.IsImmediate()) {
return fmt::format("int({})", offset.U32());
}
IR::Inst* const inst{offset.InstRecursive()};
if (inst->AreAllArgsImmediates()) {
switch (inst->GetOpcode()) {
case IR::Opcode::CompositeConstructU32x2:
return fmt::format("ivec2({},{})", inst->Arg(0).U32(), inst->Arg(1).U32());
case IR::Opcode::CompositeConstructU32x3:
return fmt::format("ivec3({},{},{})", inst->Arg(0).U32(), inst->Arg(1).U32(),
inst->Arg(2).U32());
case IR::Opcode::CompositeConstructU32x4:
return fmt::format("ivec4({},{},{},{})", inst->Arg(0).U32(), inst->Arg(1).U32(),
inst->Arg(2).U32(), inst->Arg(3).U32());
default:
break;
}
}
const bool has_var_aoffi{ctx.profile.support_gl_variable_aoffi};
if (!has_var_aoffi) {
LOG_WARNING(Shader_GLSL, "Device does not support variable texture offsets, STUBBING");
}
const auto offset_str{has_var_aoffi ? ctx.var_alloc.Consume(offset) : "0"};
switch (offset.Type()) {
case IR::Type::U32:
return fmt::format("int({})", offset_str);
case IR::Type::U32x2:
return fmt::format("ivec2({})", offset_str);
case IR::Type::U32x3:
return fmt::format("ivec3({})", offset_str);
case IR::Type::U32x4:
return fmt::format("ivec4({})", offset_str);
default:
throw NotImplementedException("Offset type {}", offset.Type());
}
}
std::string PtpOffsets(const IR::Value& offset, const IR::Value& offset2) {
const std::array values{offset.InstRecursive(), offset2.InstRecursive()};
if (!values[0]->AreAllArgsImmediates() || !values[1]->AreAllArgsImmediates()) {
LOG_WARNING(Shader_GLSL, "Not all arguments in PTP are immediate, STUBBING");
return "ivec2[](ivec2(0), ivec2(1), ivec2(2), ivec2(3))";
}
const IR::Opcode opcode{values[0]->GetOpcode()};
if (opcode != values[1]->GetOpcode() || opcode != IR::Opcode::CompositeConstructU32x4) {
throw LogicError("Invalid PTP arguments");
}
auto read{[&](unsigned int a, unsigned int b) { return values[a]->Arg(b).U32(); }};
return fmt::format("ivec2[](ivec2({},{}),ivec2({},{}),ivec2({},{}),ivec2({},{}))", read(0, 0),
read(0, 1), read(0, 2), read(0, 3), read(1, 0), read(1, 1), read(1, 2),
read(1, 3));
}
IR::Inst* PrepareSparse(IR::Inst& inst) {
const auto sparse_inst{inst.GetAssociatedPseudoOperation(IR::Opcode::GetSparseFromOp)};
if (sparse_inst) {
sparse_inst->Invalidate();
}
return sparse_inst;
}
std::string ImageGatherSubpixelOffset(const IR::TextureInstInfo& info, std::string_view texture,
std::string_view coords) {
switch (info.type) {
case TextureType::Color2D:
case TextureType::Color2DRect:
return fmt::format("{}+vec2(0.001953125)/vec2(textureSize({}, 0))", coords, texture);
case TextureType::ColorArray2D:
case TextureType::ColorCube:
return fmt::format("vec3({0}.xy+vec2(0.001953125)/vec2(textureSize({1}, 0)),{0}.z)", coords,
texture);
default:
return std::string{coords};
}
}
} // Anonymous namespace
void EmitImageSampleImplicitLod(EmitContext& ctx, IR::Inst& inst, const IR::Value& index,
std::string_view coords, std::string_view bias_lc,
const IR::Value& offset) {
const auto info{inst.Flags<IR::TextureInstInfo>()};
if (info.has_lod_clamp) {
throw NotImplementedException("EmitImageSampleImplicitLod Lod clamp samples");
}
const auto texture{Texture(ctx, info, index)};
const auto bias{info.has_bias ? fmt::format(",{}", bias_lc) : ""};
const auto texel{ctx.var_alloc.Define(inst, GlslVarType::F32x4)};
const auto sparse_inst{PrepareSparse(inst)};
const bool supports_sparse{ctx.profile.support_gl_sparse_textures};
if (sparse_inst && !supports_sparse) {
LOG_WARNING(Shader_GLSL, "Device does not support sparse texture queries. STUBBING");
ctx.AddU1("{}=true;", *sparse_inst);
}
if (!sparse_inst || !supports_sparse) {
if (!offset.IsEmpty()) {
const auto offset_str{GetOffsetVec(ctx, offset)};
if (ctx.stage == Stage::Fragment) {
ctx.Add("{}=textureOffset({},{},{}{});", texel, texture, coords, offset_str, bias);
} else {
ctx.Add("{}=textureLodOffset({},{},0.0,{});", texel, texture, coords, offset_str);
}
} else {
if (ctx.stage == Stage::Fragment) {
ctx.Add("{}=texture({},{}{});", texel, texture, coords, bias);
} else {
ctx.Add("{}=textureLod({},{},0.0);", texel, texture, coords);
}
}
return;
}
if (!offset.IsEmpty()) {
ctx.AddU1("{}=sparseTexelsResidentARB(sparseTextureOffsetARB({},{},{},{}{}));",
*sparse_inst, texture, coords, GetOffsetVec(ctx, offset), texel, bias);
} else {
ctx.AddU1("{}=sparseTexelsResidentARB(sparseTextureARB({},{},{}{}));", *sparse_inst,
texture, coords, texel, bias);
}
}
void EmitImageSampleExplicitLod(EmitContext& ctx, IR::Inst& inst, const IR::Value& index,
std::string_view coords, std::string_view lod_lc,
const IR::Value& offset) {
const auto info{inst.Flags<IR::TextureInstInfo>()};
if (info.has_bias) {
throw NotImplementedException("EmitImageSampleExplicitLod Bias texture samples");
}
if (info.has_lod_clamp) {
throw NotImplementedException("EmitImageSampleExplicitLod Lod clamp samples");
}
const auto texture{Texture(ctx, info, index)};
const auto texel{ctx.var_alloc.Define(inst, GlslVarType::F32x4)};
const auto sparse_inst{PrepareSparse(inst)};
const bool supports_sparse{ctx.profile.support_gl_sparse_textures};
if (sparse_inst && !supports_sparse) {
LOG_WARNING(Shader_GLSL, "Device does not support sparse texture queries. STUBBING");
ctx.AddU1("{}=true;", *sparse_inst);
}
if (!sparse_inst || !supports_sparse) {
if (!offset.IsEmpty()) {
ctx.Add("{}=textureLodOffset({},{},{},{});", texel, texture, coords, lod_lc,
GetOffsetVec(ctx, offset));
} else {
ctx.Add("{}=textureLod({},{},{});", texel, texture, coords, lod_lc);
}
return;
}
if (!offset.IsEmpty()) {
ctx.AddU1("{}=sparseTexelsResidentARB(sparseTexelFetchOffsetARB({},{},int({}),{},{}));",
*sparse_inst, texture, CastToIntVec(coords, info), lod_lc,
GetOffsetVec(ctx, offset), texel);
} else {
ctx.AddU1("{}=sparseTexelsResidentARB(sparseTextureLodARB({},{},{},{}));", *sparse_inst,
texture, coords, lod_lc, texel);
}
}
void EmitImageSampleDrefImplicitLod(EmitContext& ctx, IR::Inst& inst, const IR::Value& index,
std::string_view coords, std::string_view dref,
std::string_view bias_lc, const IR::Value& offset) {
const auto info{inst.Flags<IR::TextureInstInfo>()};
const auto sparse_inst{PrepareSparse(inst)};
if (sparse_inst) {
throw NotImplementedException("EmitImageSampleDrefImplicitLod Sparse texture samples");
}
if (info.has_bias) {
throw NotImplementedException("EmitImageSampleDrefImplicitLod Bias texture samples");
}
if (info.has_lod_clamp) {
throw NotImplementedException("EmitImageSampleDrefImplicitLod Lod clamp samples");
}
const auto texture{Texture(ctx, info, index)};
const auto bias{info.has_bias ? fmt::format(",{}", bias_lc) : ""};
const bool needs_shadow_ext{NeedsShadowLodExt(info.type)};
const auto cast{needs_shadow_ext ? "vec4" : "vec3"};
const bool use_grad{!ctx.profile.support_gl_texture_shadow_lod &&
ctx.stage != Stage::Fragment && needs_shadow_ext};
if (use_grad) {
LOG_WARNING(Shader_GLSL,
"Device lacks GL_EXT_texture_shadow_lod. Using textureGrad fallback");
if (info.type == TextureType::ColorArrayCube) {
LOG_WARNING(Shader_GLSL, "textureGrad does not support ColorArrayCube. Stubbing");
ctx.AddF32("{}=0.0f;", inst);
return;
}
const auto d_cast{info.type == TextureType::ColorArray2D ? "vec2" : "vec3"};
ctx.AddF32("{}=textureGrad({},{}({},{}),{}(0),{}(0));", inst, texture, cast, coords, dref,
d_cast, d_cast);
return;
}
if (!offset.IsEmpty()) {
const auto offset_str{GetOffsetVec(ctx, offset)};
if (ctx.stage == Stage::Fragment) {
ctx.AddF32("{}=textureOffset({},{}({},{}),{}{});", inst, texture, cast, coords, dref,
offset_str, bias);
} else {
ctx.AddF32("{}=textureLodOffset({},{}({},{}),0.0,{});", inst, texture, cast, coords,
dref, offset_str);
}
} else {
if (ctx.stage == Stage::Fragment) {
if (info.type == TextureType::ColorArrayCube) {
ctx.AddF32("{}=texture({},vec4({}),{});", inst, texture, coords, dref);
} else {
ctx.AddF32("{}=texture({},{}({},{}){});", inst, texture, cast, coords, dref, bias);
}
} else {
ctx.AddF32("{}=textureLod({},{}({},{}),0.0);", inst, texture, cast, coords, dref);
}
}
}
void EmitImageSampleDrefExplicitLod(EmitContext& ctx, IR::Inst& inst, const IR::Value& index,
std::string_view coords, std::string_view dref,
std::string_view lod_lc, const IR::Value& offset) {
const auto info{inst.Flags<IR::TextureInstInfo>()};
const auto sparse_inst{PrepareSparse(inst)};
if (sparse_inst) {
throw NotImplementedException("EmitImageSampleDrefExplicitLod Sparse texture samples");
}
if (info.has_bias) {
throw NotImplementedException("EmitImageSampleDrefExplicitLod Bias texture samples");
}
if (info.has_lod_clamp) {
throw NotImplementedException("EmitImageSampleDrefExplicitLod Lod clamp samples");
}
const auto texture{Texture(ctx, info, index)};
const bool needs_shadow_ext{NeedsShadowLodExt(info.type)};
const bool use_grad{!ctx.profile.support_gl_texture_shadow_lod && needs_shadow_ext};
const auto cast{needs_shadow_ext ? "vec4" : "vec3"};
if (use_grad) {
LOG_WARNING(Shader_GLSL,
"Device lacks GL_EXT_texture_shadow_lod. Using textureGrad fallback");
if (info.type == TextureType::ColorArrayCube) {
LOG_WARNING(Shader_GLSL, "textureGrad does not support ColorArrayCube. Stubbing");
ctx.AddF32("{}=0.0f;", inst);
return;
}
const auto d_cast{info.type == TextureType::ColorArray2D ? "vec2" : "vec3"};
ctx.AddF32("{}=textureGrad({},{}({},{}),{}(0),{}(0));", inst, texture, cast, coords, dref,
d_cast, d_cast);
return;
}
if (!offset.IsEmpty()) {
const auto offset_str{GetOffsetVec(ctx, offset)};
if (info.type == TextureType::ColorArrayCube) {
ctx.AddF32("{}=textureLodOffset({},{},{},{},{});", inst, texture, coords, dref, lod_lc,
offset_str);
} else {
ctx.AddF32("{}=textureLodOffset({},{}({},{}),{},{});", inst, texture, cast, coords,
dref, lod_lc, offset_str);
}
} else {
if (info.type == TextureType::ColorArrayCube) {
ctx.AddF32("{}=textureLod({},{},{},{});", inst, texture, coords, dref, lod_lc);
} else {
ctx.AddF32("{}=textureLod({},{}({},{}),{});", inst, texture, cast, coords, dref,
lod_lc);
}
}
}
void EmitImageGather(EmitContext& ctx, IR::Inst& inst, const IR::Value& index,
std::string_view coords, const IR::Value& offset, const IR::Value& offset2) {
const auto info{inst.Flags<IR::TextureInstInfo>()};
const auto texture{Texture(ctx, info, index)};
const auto texel{ctx.var_alloc.Define(inst, GlslVarType::F32x4)};
const auto sparse_inst{PrepareSparse(inst)};
const bool supports_sparse{ctx.profile.support_gl_sparse_textures};
if (sparse_inst && !supports_sparse) {
LOG_WARNING(Shader_GLSL, "Device does not support sparse texture queries. STUBBING");
ctx.AddU1("{}=true;", *sparse_inst);
}
std::string coords_with_subpixel_offset;
if (ctx.profile.need_gather_subpixel_offset) {
// Apply a subpixel offset of 1/512 the texel size of the texture to ensure same rounding on
// AMD hardware as on Maxwell or other Nvidia architectures.
coords_with_subpixel_offset = ImageGatherSubpixelOffset(info, texture, coords);
coords = coords_with_subpixel_offset;
}
if (!sparse_inst || !supports_sparse) {
if (offset.IsEmpty()) {
ctx.Add("{}=textureGather({},{},int({}));", texel, texture, coords,
info.gather_component);
return;
}
if (offset2.IsEmpty()) {
ctx.Add("{}=textureGatherOffset({},{},{},int({}));", texel, texture, coords,
GetOffsetVec(ctx, offset), info.gather_component);
return;
}
// PTP
const auto offsets{PtpOffsets(offset, offset2)};
ctx.Add("{}=textureGatherOffsets({},{},{},int({}));", texel, texture, coords, offsets,
info.gather_component);
return;
}
if (offset.IsEmpty()) {
ctx.AddU1("{}=sparseTexelsResidentARB(sparseTextureGatherARB({},{},{},int({})));",
*sparse_inst, texture, coords, texel, info.gather_component);
return;
}
if (offset2.IsEmpty()) {
ctx.AddU1("{}=sparseTexelsResidentARB(sparseTextureGatherOffsetARB({},{},{},{},int({})));",
*sparse_inst, texture, CastToIntVec(coords, info), GetOffsetVec(ctx, offset),
texel, info.gather_component);
return;
}
// PTP
const auto offsets{PtpOffsets(offset, offset2)};
ctx.AddU1("{}=sparseTexelsResidentARB(sparseTextureGatherOffsetARB({},{},{},{},int({})));",
*sparse_inst, texture, CastToIntVec(coords, info), offsets, texel,
info.gather_component);
}
void EmitImageGatherDref(EmitContext& ctx, IR::Inst& inst, const IR::Value& index,
std::string_view coords, const IR::Value& offset, const IR::Value& offset2,
std::string_view dref) {
const auto info{inst.Flags<IR::TextureInstInfo>()};
const auto texture{Texture(ctx, info, index)};
const auto texel{ctx.var_alloc.Define(inst, GlslVarType::F32x4)};
const auto sparse_inst{PrepareSparse(inst)};
const bool supports_sparse{ctx.profile.support_gl_sparse_textures};
if (sparse_inst && !supports_sparse) {
LOG_WARNING(Shader_GLSL, "Device does not support sparse texture queries. STUBBING");
ctx.AddU1("{}=true;", *sparse_inst);
}
std::string coords_with_subpixel_offset;
if (ctx.profile.need_gather_subpixel_offset) {
// Apply a subpixel offset of 1/512 the texel size of the texture to ensure same rounding on
// AMD hardware as on Maxwell or other Nvidia architectures.
coords_with_subpixel_offset = ImageGatherSubpixelOffset(info, texture, coords);
coords = coords_with_subpixel_offset;
}
if (!sparse_inst || !supports_sparse) {
if (offset.IsEmpty()) {
ctx.Add("{}=textureGather({},{},{});", texel, texture, coords, dref);
return;
}
if (offset2.IsEmpty()) {
ctx.Add("{}=textureGatherOffset({},{},{},{});", texel, texture, coords, dref,
GetOffsetVec(ctx, offset));
return;
}
// PTP
const auto offsets{PtpOffsets(offset, offset2)};
ctx.Add("{}=textureGatherOffsets({},{},{},{});", texel, texture, coords, dref, offsets);
return;
}
if (offset.IsEmpty()) {
ctx.AddU1("{}=sparseTexelsResidentARB(sparseTextureGatherARB({},{},{},{}));", *sparse_inst,
texture, coords, dref, texel);
return;
}
if (offset2.IsEmpty()) {
ctx.AddU1("{}=sparseTexelsResidentARB(sparseTextureGatherOffsetARB({},{},{},,{},{}));",
*sparse_inst, texture, CastToIntVec(coords, info), dref,
GetOffsetVec(ctx, offset), texel);
return;
}
// PTP
const auto offsets{PtpOffsets(offset, offset2)};
ctx.AddU1("{}=sparseTexelsResidentARB(sparseTextureGatherOffsetARB({},{},{},,{},{}));",
*sparse_inst, texture, CastToIntVec(coords, info), dref, offsets, texel);
}
void EmitImageFetch(EmitContext& ctx, IR::Inst& inst, const IR::Value& index,
std::string_view coords, std::string_view offset, std::string_view lod,
std::string_view ms) {
const auto info{inst.Flags<IR::TextureInstInfo>()};
if (info.has_bias) {
throw NotImplementedException("EmitImageFetch Bias texture samples");
}
if (info.has_lod_clamp) {
throw NotImplementedException("EmitImageFetch Lod clamp samples");
}
const auto texture{Texture(ctx, info, index)};
const auto sparse_inst{PrepareSparse(inst)};
const auto texel{ctx.var_alloc.Define(inst, GlslVarType::F32x4)};
const bool supports_sparse{ctx.profile.support_gl_sparse_textures};
if (sparse_inst && !supports_sparse) {
LOG_WARNING(Shader_GLSL, "Device does not support sparse texture queries. STUBBING");
ctx.AddU1("{}=true;", *sparse_inst);
}
if (!sparse_inst || !supports_sparse) {
const auto int_coords{CoordsCastToInt(coords, info)};
if (!ms.empty()) {
ctx.Add("{}=texelFetch({},{},int({}));", texel, texture, int_coords, ms);
} else if (!offset.empty()) {
ctx.Add("{}=texelFetchOffset({},{},int({}),{});", texel, texture, int_coords, lod,
CoordsCastToInt(offset, info));
} else {
if (info.type == TextureType::Buffer) {
ctx.Add("{}=texelFetch({},int({}));", texel, texture, coords);
} else {
ctx.Add("{}=texelFetch({},{},int({}));", texel, texture, int_coords, lod);
}
}
return;
}
if (!ms.empty()) {
throw NotImplementedException("EmitImageFetch Sparse MSAA samples");
}
if (!offset.empty()) {
ctx.AddU1("{}=sparseTexelsResidentARB(sparseTexelFetchOffsetARB({},{},int({}),{},{}));",
*sparse_inst, texture, CastToIntVec(coords, info), lod,
CastToIntVec(offset, info), texel);
} else {
ctx.AddU1("{}=sparseTexelsResidentARB(sparseTexelFetchARB({},{},int({}),{}));",
*sparse_inst, texture, CastToIntVec(coords, info), lod, texel);
}
}
void EmitImageQueryDimensions(EmitContext& ctx, IR::Inst& inst, const IR::Value& index,
std::string_view lod, const IR::Value& skip_mips_val) {
const auto info{inst.Flags<IR::TextureInstInfo>()};
const auto texture{Texture(ctx, info, index)};
const bool is_msaa{IsTextureMsaa(ctx, info)};
const bool skip_mips{skip_mips_val.U1()};
const auto mips{skip_mips ? "0u" : fmt::format("uint(textureQueryLevels({}))", texture)};
if (is_msaa && !skip_mips) {
throw NotImplementedException("EmitImageQueryDimensions MSAA QueryLevels");
}
if (info.type == TextureType::Buffer && !skip_mips) {
throw NotImplementedException("EmitImageQueryDimensions TextureType::Buffer QueryLevels");
}
const bool uses_lod{!is_msaa && info.type != TextureType::Buffer};
const auto lod_str{uses_lod ? fmt::format(",int({})", lod) : ""};
switch (info.type) {
case TextureType::Color1D:
return ctx.AddU32x4("{}=uvec4(uint(textureSize({}{})),0u,0u,{});", inst, texture, lod_str,
mips);
case TextureType::ColorArray1D:
case TextureType::Color2D:
case TextureType::ColorCube:
case TextureType::Color2DRect:
return ctx.AddU32x4("{}=uvec4(uvec2(textureSize({}{})),0u,{});", inst, texture, lod_str,
mips);
case TextureType::ColorArray2D:
case TextureType::Color3D:
case TextureType::ColorArrayCube:
return ctx.AddU32x4("{}=uvec4(uvec3(textureSize({}{})),{});", inst, texture, lod_str, mips);
case TextureType::Buffer:
return ctx.AddU32x4("{}=uvec4(uint(textureSize({})),0u,0u,{});", inst, texture, mips);
}
throw LogicError("Unspecified image type {}", info.type.Value());
}
void EmitImageQueryLod(EmitContext& ctx, IR::Inst& inst, const IR::Value& index,
std::string_view coords) {
const auto info{inst.Flags<IR::TextureInstInfo>()};
const auto texture{Texture(ctx, info, index)};
return ctx.AddF32x4("{}=vec4(textureQueryLod({},{}),0.0,0.0);", inst, texture, coords);
}
void EmitImageGradient(EmitContext& ctx, IR::Inst& inst, const IR::Value& index,
std::string_view coords, const IR::Value& derivatives,
const IR::Value& offset, [[maybe_unused]] const IR::Value& lod_clamp) {
const auto info{inst.Flags<IR::TextureInstInfo>()};
if (info.has_lod_clamp) {
throw NotImplementedException("EmitImageGradient Lod clamp samples");
}
const auto sparse_inst{PrepareSparse(inst)};
if (sparse_inst) {
throw NotImplementedException("EmitImageGradient Sparse");
}
if (!offset.IsEmpty() && info.num_derivates <= 2) {
throw NotImplementedException("EmitImageGradient offset");
}
const auto texture{Texture(ctx, info, index)};
const auto texel{ctx.var_alloc.Define(inst, GlslVarType::F32x4)};
const bool multi_component{info.num_derivates > 1 || info.has_lod_clamp};
const auto derivatives_vec{ctx.var_alloc.Consume(derivatives)};
if (multi_component) {
if (info.num_derivates >= 3) {
const auto offset_vec{ctx.var_alloc.Consume(offset)};
ctx.Add("{}=textureGrad({},{},vec3({}.xz, {}.x),vec3({}.yz, {}.y));", texel, texture,
coords, derivatives_vec, offset_vec, derivatives_vec, offset_vec);
return;
}
ctx.Add("{}=textureGrad({},{},vec2({}.xz),vec2({}.yz));", texel, texture, coords,
derivatives_vec, derivatives_vec);
} else {
ctx.Add("{}=textureGrad({},{},float({}.x),float({}.y));", texel, texture, coords,
derivatives_vec, derivatives_vec);
}
}
void EmitImageRead(EmitContext& ctx, IR::Inst& inst, const IR::Value& index,
std::string_view coords) {
const auto info{inst.Flags<IR::TextureInstInfo>()};
const auto sparse_inst{PrepareSparse(inst)};
if (sparse_inst) {
throw NotImplementedException("EmitImageRead Sparse");
}
const auto image{Image(ctx, info, index)};
ctx.AddU32x4("{}=uvec4(imageLoad({},{}));", inst, image, CoordsCastToInt(coords, info));
}
void EmitImageWrite(EmitContext& ctx, IR::Inst& inst, const IR::Value& index,
std::string_view coords, std::string_view color) {
const auto info{inst.Flags<IR::TextureInstInfo>()};
const auto image{Image(ctx, info, index)};
ctx.Add("imageStore({},{},{});", image, CoordsCastToInt(coords, info), color);
}
void EmitImageAtomicIAdd32(EmitContext& ctx, IR::Inst& inst, const IR::Value& index,
std::string_view coords, std::string_view value) {
const auto info{inst.Flags<IR::TextureInstInfo>()};
const auto image{Image(ctx, info, index)};
ctx.AddU32("{}=imageAtomicAdd({},{},{});", inst, image, CoordsCastToInt(coords, info), value);
}
void EmitImageAtomicSMin32(EmitContext& ctx, IR::Inst& inst, const IR::Value& index,
std::string_view coords, std::string_view value) {
const auto info{inst.Flags<IR::TextureInstInfo>()};
const auto image{Image(ctx, info, index)};
ctx.AddU32("{}=imageAtomicMin({},{},int({}));", inst, image, CoordsCastToInt(coords, info),
value);
}
void EmitImageAtomicUMin32(EmitContext& ctx, IR::Inst& inst, const IR::Value& index,
std::string_view coords, std::string_view value) {
const auto info{inst.Flags<IR::TextureInstInfo>()};
const auto image{Image(ctx, info, index)};
ctx.AddU32("{}=imageAtomicMin({},{},uint({}));", inst, image, CoordsCastToInt(coords, info),
value);
}
void EmitImageAtomicSMax32(EmitContext& ctx, IR::Inst& inst, const IR::Value& index,
std::string_view coords, std::string_view value) {
const auto info{inst.Flags<IR::TextureInstInfo>()};
const auto image{Image(ctx, info, index)};
ctx.AddU32("{}=imageAtomicMax({},{},int({}));", inst, image, CoordsCastToInt(coords, info),
value);
}
void EmitImageAtomicUMax32(EmitContext& ctx, IR::Inst& inst, const IR::Value& index,
std::string_view coords, std::string_view value) {
const auto info{inst.Flags<IR::TextureInstInfo>()};
const auto image{Image(ctx, info, index)};
ctx.AddU32("{}=imageAtomicMax({},{},uint({}));", inst, image, CoordsCastToInt(coords, info),
value);
}
void EmitImageAtomicInc32(EmitContext&, IR::Inst&, const IR::Value&, std::string_view,
std::string_view) {
NotImplemented();
}
void EmitImageAtomicDec32(EmitContext&, IR::Inst&, const IR::Value&, std::string_view,
std::string_view) {
NotImplemented();
}
void EmitImageAtomicAnd32(EmitContext& ctx, IR::Inst& inst, const IR::Value& index,
std::string_view coords, std::string_view value) {
const auto info{inst.Flags<IR::TextureInstInfo>()};
const auto image{Image(ctx, info, index)};
ctx.AddU32("{}=imageAtomicAnd({},{},{});", inst, image, CoordsCastToInt(coords, info), value);
}
void EmitImageAtomicOr32(EmitContext& ctx, IR::Inst& inst, const IR::Value& index,
std::string_view coords, std::string_view value) {
const auto info{inst.Flags<IR::TextureInstInfo>()};
const auto image{Image(ctx, info, index)};
ctx.AddU32("{}=imageAtomicOr({},{},{});", inst, image, CoordsCastToInt(coords, info), value);
}
void EmitImageAtomicXor32(EmitContext& ctx, IR::Inst& inst, const IR::Value& index,
std::string_view coords, std::string_view value) {
const auto info{inst.Flags<IR::TextureInstInfo>()};
const auto image{Image(ctx, info, index)};
ctx.AddU32("{}=imageAtomicXor({},{},{});", inst, image, CoordsCastToInt(coords, info), value);
}
void EmitImageAtomicExchange32(EmitContext& ctx, IR::Inst& inst, const IR::Value& index,
std::string_view coords, std::string_view value) {
const auto info{inst.Flags<IR::TextureInstInfo>()};
const auto image{Image(ctx, info, index)};
ctx.AddU32("{}=imageAtomicExchange({},{},{});", inst, image, CoordsCastToInt(coords, info),
value);
}
void EmitIsTextureScaled(EmitContext& ctx, IR::Inst& inst, const IR::Value& index) {
if (!index.IsImmediate()) {
throw NotImplementedException("Non-constant texture rescaling");
}
const u32 image_index{index.U32()};
ctx.AddU1("{}=(ftou(scaling.x)&{})!=0;", inst, 1u << image_index);
}
void EmitIsImageScaled(EmitContext& ctx, IR::Inst& inst, const IR::Value& index) {
if (!index.IsImmediate()) {
throw NotImplementedException("Non-constant texture rescaling");
}
const u32 image_index{index.U32()};
ctx.AddU1("{}=(ftou(scaling.y)&{})!=0;", inst, 1u << image_index);
}
void EmitBindlessImageSampleImplicitLod(EmitContext&) {
NotImplemented();
}
void EmitBindlessImageSampleExplicitLod(EmitContext&) {
NotImplemented();
}
void EmitBindlessImageSampleDrefImplicitLod(EmitContext&) {
NotImplemented();
}
void EmitBindlessImageSampleDrefExplicitLod(EmitContext&) {
NotImplemented();
}
void EmitBindlessImageGather(EmitContext&) {
NotImplemented();
}
void EmitBindlessImageGatherDref(EmitContext&) {
NotImplemented();
}
void EmitBindlessImageFetch(EmitContext&) {
NotImplemented();
}
void EmitBindlessImageQueryDimensions(EmitContext&) {
NotImplemented();
}
void EmitBindlessImageQueryLod(EmitContext&) {
NotImplemented();
}
void EmitBindlessImageGradient(EmitContext&) {
NotImplemented();
}
void EmitBindlessImageRead(EmitContext&) {
NotImplemented();
}
void EmitBindlessImageWrite(EmitContext&) {
NotImplemented();
}
void EmitBoundImageSampleImplicitLod(EmitContext&) {
NotImplemented();
}
void EmitBoundImageSampleExplicitLod(EmitContext&) {
NotImplemented();
}
void EmitBoundImageSampleDrefImplicitLod(EmitContext&) {
NotImplemented();
}
void EmitBoundImageSampleDrefExplicitLod(EmitContext&) {
NotImplemented();
}
void EmitBoundImageGather(EmitContext&) {
NotImplemented();
}
void EmitBoundImageGatherDref(EmitContext&) {
NotImplemented();
}
void EmitBoundImageFetch(EmitContext&) {
NotImplemented();
}
void EmitBoundImageQueryDimensions(EmitContext&) {
NotImplemented();
}
void EmitBoundImageQueryLod(EmitContext&) {
NotImplemented();
}
void EmitBoundImageGradient(EmitContext&) {
NotImplemented();
}
void EmitBoundImageRead(EmitContext&) {
NotImplemented();
}
void EmitBoundImageWrite(EmitContext&) {
NotImplemented();
}
void EmitBindlessImageAtomicIAdd32(EmitContext&) {
NotImplemented();
}
void EmitBindlessImageAtomicSMin32(EmitContext&) {
NotImplemented();
}
void EmitBindlessImageAtomicUMin32(EmitContext&) {
NotImplemented();
}
void EmitBindlessImageAtomicSMax32(EmitContext&) {
NotImplemented();
}
void EmitBindlessImageAtomicUMax32(EmitContext&) {
NotImplemented();
}
void EmitBindlessImageAtomicInc32(EmitContext&) {
NotImplemented();
}
void EmitBindlessImageAtomicDec32(EmitContext&) {
NotImplemented();
}
void EmitBindlessImageAtomicAnd32(EmitContext&) {
NotImplemented();
}
void EmitBindlessImageAtomicOr32(EmitContext&) {
NotImplemented();
}
void EmitBindlessImageAtomicXor32(EmitContext&) {
NotImplemented();
}
void EmitBindlessImageAtomicExchange32(EmitContext&) {
NotImplemented();
}
void EmitBoundImageAtomicIAdd32(EmitContext&) {
NotImplemented();
}
void EmitBoundImageAtomicSMin32(EmitContext&) {
NotImplemented();
}
void EmitBoundImageAtomicUMin32(EmitContext&) {
NotImplemented();
}
void EmitBoundImageAtomicSMax32(EmitContext&) {
NotImplemented();
}
void EmitBoundImageAtomicUMax32(EmitContext&) {
NotImplemented();
}
void EmitBoundImageAtomicInc32(EmitContext&) {
NotImplemented();
}
void EmitBoundImageAtomicDec32(EmitContext&) {
NotImplemented();
}
void EmitBoundImageAtomicAnd32(EmitContext&) {
NotImplemented();
}
void EmitBoundImageAtomicOr32(EmitContext&) {
NotImplemented();
}
void EmitBoundImageAtomicXor32(EmitContext&) {
NotImplemented();
}
void EmitBoundImageAtomicExchange32(EmitContext&) {
NotImplemented();
}
} // namespace Shader::Backend::GLSL
|