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
use *;
/// Defines how new pixels are composited with existing pixels on the canvas.
///
/// Maps directly to the CSS `globalCompositeOperation` property.
/// Rendering quality preset controlling anti-aliasing smoothing strategy.
///
/// Maps to the canvas `imageSmoothingQuality` value plus an explicit
/// `imageSmoothingEnabled` toggle. Combined with a CSS `image-rendering:
/// pixelated` rule on the consumer side, `Low` produces crisp pixel-art
/// rendering while `High` produces smooth vector-style rendering.
/// Whether a vertex buffer is consumed per-vertex or per-instance.
/// What a render pass attachment does with the contents it already holds
/// before the pass draws anything.
///
/// Replaces the `"clear"` / `"load"` string literals that
/// `GpuRenderPassColorAttachment.loadOp` and
/// `GpuRenderPassDepthStencilAttachment.depthLoadOp` accept. A typo in a
/// string literal is a silent runtime failure (the browser rejects the pass
/// and the frame renders nothing); a typo in a variant is a compile error.
/// What a render pass attachment does with the contents it produced once the
/// pass ends.
///
/// Replaces the `"store"` / `"discard"` string literals accepted by
/// `storeOp` and `depthStoreOp`. `Discard` saves bandwidth for attachments
/// the pass writes but nothing ever reads back — the depth buffer of a
/// depth-test-only pass is the canonical case.
/// One legal use of a GPU buffer, as declared in `GpuBufferDescriptor.usage`.
///
/// The WebGPU buffer usage field is a bitmask, not an enum: a buffer may
/// be used for several purposes at once. This enum names the individual
/// bits so call sites never hand-write a magic number, and the renderer
/// ORs the selected variants into the bitmask at the call site. Combine
/// bits with `|`, e.g. `BufferUsage::Vertex | BufferUsage::CopyDestination`.
/// One legal use of a GPU texture, as declared in `GpuTextureDescriptor.usage`.
///
/// Like [`BufferUsage`], the underlying WebGPU field is a bitmask and the
/// renderer ORs the selected variants together at the call site. A render
/// target needs `RenderAttachment`; a sampled texture needs
/// `TextureBinding`; a write-once compute output needs `StorageBinding`.
/// How texels are interpolated when a sample is fetched.
///
/// Used for both the magnification filter (`magFilter`) and the
/// minification filter (`minFilter`) of a sampler. The two filters are
/// separate WebGPU fields because a magnified surface and a minified
/// surface of the same texture want different behaviour — pixel art wants
/// `Nearest` when magnified and `Linear` when minified — so a single
/// [`SamplerDescriptor::filter`] covers both by letting the caller pick
/// the same mode or override it per stage.
/// How texels are blended between adjacent mip levels of a mipmapped
/// texture.
///
/// This is the `mipmapFilter` field of a sampler, which is a separate
/// decision from [`FilterMode`]: `FilterMode` decides how one mip level is
/// sampled, this decides how the results of two levels are combined.
/// What a sampler does with coordinates that fall outside `[0, 1]`.
///
/// One value per texture axis (`addressModeU`, `addressModeV`,
/// `addressModeW` on the wire). The W axis is only meaningful for 3D
/// textures; for 2D textures it is accepted and ignored.
/// The comparison performed between two values, used both for depth tests
/// and for comparison samplers.
///
/// Replaces the bare `&'static str` that
/// `GpuSamplerDescriptor.compare` could not express: a `bool` there forced
/// the renderer to hardcode a single comparison, while
/// [`SamplerDescriptor::compare`] carries the comparison itself as
/// `Option<CompareFunction>` — `None` for a plain filtering sampler.
/// One term of a blend equation: the factor the source or destination value
/// is multiplied by.
///
/// A blend state pairs one of these as the source factor and one as the
/// destination factor, then combines the two results with a
/// [`BlendOperation`]. The conventional alpha-blend pairing is
/// `SourceAlpha` as the source and `OneMinusSourceAlpha` as the
/// destination, with `BlendOperation::Add`.
/// How the two blended terms are combined into the final value.
/// How the vertex stage assembles the stream of vertices into primitives.
/// The element width of the index buffer handed to `setIndexBuffer`.
///
/// The format must match the element type of the index data and the
/// [`PrimitiveState::strip_index_format`] when a strip topology is used.
/// The shader stages a bind group layout entry is visible to.
///
/// The wire field is a bitmask, so entries are usually visible to more
/// than one stage; the renderer ORs the selected variants together at the
/// call site. This replaces the raw `visibility: u32` on
/// [`BindGroupLayoutEntry`], where `0x1` / `0x2` / `0x4` were written by
/// hand and a wrong bit silently produced a binding the shader cannot see.
/// Which winding order counts as the front face of a triangle.
///
/// The two sides are labelled only relative to this choice, so flipping
/// it swaps what [`CullMode::Front`] and [`CullMode::Back`] discard.
/// Which triangle faces are discarded before rasterization.
/// The in-memory layout of one vertex attribute.
///
/// Maps to the `GPUVertexAttribute.format` field. A position of three
/// `f32` is `Float32x3`; a packed 8-bit-per-channel color is `Unorm8x4`.
/// The texture formats the renderer supports, as a closed set.
///
/// Deliberately a subset of the WebGPU format list: every variant here is
/// one the renderer's attachment, binding, and sampling paths are known
/// to handle. Naming them as an enum turns a wrong format string into a
/// compile error rather than a validation-layer rejection at creation time.
/// The class of error a WebGPU error scope collects.
///
/// Pushed onto `GpuDevice` via
/// [`WebGpuRenderer::push_error_scope`]; every error of this class raised
/// while the scope is open is captured instead of being reported to the
/// console immediately. A typo in the filter string is a silent
/// behaviour change - `pushErrorScope` accepts any string and simply
/// never matches - so the filter is named as a variant.