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
use wgpu_upstream::{BufferBinding, SamplerBindingType};
use crate as wgpu;
/// A type aimed at simplifying the creation of a bind group layout.
#[derive(Debug, Default)]
pub struct LayoutBuilder {
bindings: Vec<(wgpu::ShaderStages, wgpu::BindingType)>,
}
/// Simplified creation of a bind group.
#[derive(Debug, Default)]
pub struct Builder<'a> {
resources: Vec<wgpu::BindingResource<'a>>,
}
impl LayoutBuilder {
/// Begin building the bind group layout.
pub fn new() -> Self {
Self::default()
}
/// Specify a new binding.
///
/// The `binding` position of each binding will be inferred as the index within the order that
/// they are added to this builder type. If you require manually specifying the binding
/// location, you may be better off not using the `BindGroupLayoutBuilder` and instead
/// constructing the `BindGroupLayout` and `BindGroup` manually.
pub fn binding(mut self, visibility: wgpu::ShaderStages, ty: wgpu::BindingType) -> Self {
self.bindings.push((visibility, ty));
self
}
/// Add a uniform buffer binding to the layout.
pub fn uniform_buffer(self, visibility: wgpu::ShaderStages, has_dynamic_offset: bool) -> Self {
let ty = wgpu::BindingType::Buffer {
ty: wgpu::BufferBindingType::Uniform,
has_dynamic_offset,
// wgpu 0.5-0.6 TODO: potential perf hit, investigate this field
min_binding_size: None,
};
self.binding(visibility, ty)
}
/// Add a storage buffer binding to the layout.
pub fn storage_buffer(
self,
visibility: wgpu::ShaderStages,
has_dynamic_offset: bool,
read_only: bool,
) -> Self {
let ty = wgpu::BindingType::Buffer {
ty: wgpu::BufferBindingType::Storage { read_only },
has_dynamic_offset,
// wgpu 0.5-0.6 TODO: potential perf hit, investigate this field
min_binding_size: None,
};
self.binding(visibility, ty)
}
/// Add a sampler binding to the layout.
pub fn sampler(self, visibility: wgpu::ShaderStages, filtering: bool) -> Self {
let ty = wgpu::BindingType::Sampler(if filtering {
SamplerBindingType::Filtering
} else {
SamplerBindingType::NonFiltering
});
self.binding(visibility, ty)
}
/// Add a sampler binding to the layout.
pub fn comparison_sampler(self, visibility: wgpu::ShaderStages) -> Self {
let ty = wgpu::BindingType::Sampler(SamplerBindingType::Comparison);
self.binding(visibility, ty)
}
/// Add a texture binding to the layout.
pub fn texture(
self,
visibility: wgpu::ShaderStages,
multisampled: bool,
view_dimension: wgpu::TextureViewDimension,
sample_type: wgpu::TextureSampleType,
) -> Self {
// fix sample type in certain scenarios (constraint given by wgpu)
let sample_type = if multisampled
&& matches!(
sample_type,
wgpu::TextureSampleType::Float { filterable: true }
) {
wgpu::TextureSampleType::Float { filterable: false }
} else {
sample_type
};
let ty = wgpu::BindingType::Texture {
multisampled,
view_dimension,
sample_type,
};
self.binding(visibility, ty)
}
/// Short-hand for adding a texture binding for a full view of the given texture to the layout.
///
/// The `multisampled` and `dimension` parameters are retrieved from the `Texture` itself.
///
/// Note that if you wish to take a `Cube` or `CubeArray` view of the given texture, you will
/// need to manually specify the `TextureViewDimension` via the `sampled_texture` method
/// instead.
pub fn texture_from(self, visibility: wgpu::ShaderStages, texture: &wgpu::Texture) -> Self {
self.texture(
visibility,
texture.sample_count() > 1,
texture.view_dimension(),
texture.sample_type(),
)
}
/// Add a storage texture binding to the layout.
pub fn storage_texture(
self,
visibility: wgpu::ShaderStages,
format: wgpu::TextureFormat,
view_dimension: wgpu::TextureViewDimension,
access: wgpu::StorageTextureAccess,
) -> Self {
let ty = wgpu::BindingType::StorageTexture {
view_dimension,
format,
access,
};
self.binding(visibility, ty)
}
/// Short-hand for adding a storage texture binding for a full view of the given texture to the
/// layout.
///
/// The `format`, `dimension` and `sample_type` are inferred from the given `texture`.
pub fn storage_texture_from(
self,
visibility: wgpu::ShaderStages,
texture: &wgpu::Texture,
access: wgpu::StorageTextureAccess,
) -> Self {
self.storage_texture(
visibility,
texture.format(),
texture.view_dimension(),
access,
)
}
/// Build the bind group layout from the specified parameters.
pub fn build(self, device: &wgpu::Device) -> wgpu::BindGroupLayout {
let mut entries = Vec::with_capacity(self.bindings.len());
for (i, (visibility, ty)) in self.bindings.into_iter().enumerate() {
let layout_binding = wgpu::BindGroupLayoutEntry {
binding: i as u32,
visibility,
ty,
count: None,
};
entries.push(layout_binding);
}
let descriptor = wgpu::BindGroupLayoutDescriptor {
label: Some("nannou bind group layout"),
entries: &entries,
};
device.create_bind_group_layout(&descriptor)
}
}
impl<'a> Builder<'a> {
/// Begin building the bind group.
pub fn new() -> Self {
Self::default()
}
/// Specify a new binding.
///
/// The `binding` position of each binding will be inferred as the index within the order that
/// they are added to this builder type. If you require manually specifying the binding
/// location, you may be better off not using the `BindGroupBuilder` and instead constructing
/// the `BindGroupLayout` and `BindGroup` manually.
pub fn binding(mut self, resource: wgpu::BindingResource<'a>) -> Self {
self.resources.push(resource);
self
}
/// Specify a slice of a buffer to be bound.
///
/// The given `range` represents the start and end point of the buffer to be bound in bytes.
pub fn buffer_bytes(
self,
buffer: &'a wgpu::Buffer,
offset: wgpu::BufferAddress,
size: Option<wgpu::BufferSize>,
) -> Self {
let resource = wgpu::BindingResource::Buffer(BufferBinding {
buffer,
offset,
size,
});
self.binding(resource)
}
/// Specify a slice of a buffer of elements of type `T` to be bound.
///
/// This method is similar to `buffer_bytes`, but expects a range of **elements** rather than a
/// range of **bytes**.
///
/// Type `T` *must* be either `#[repr(C)]` or `#[repr(transparent)]`.
// NOTE: We might want to change this to match the wgpu API by using a NonZeroU64 for size.
pub fn buffer<T>(self, buffer: &'a wgpu::Buffer, range: std::ops::Range<usize>) -> Self
where
T: Copy,
{
let size_bytes = std::mem::size_of::<T>() as wgpu::BufferAddress;
let start = range.start as wgpu::BufferAddress * size_bytes;
let end = range.end as wgpu::BufferAddress * size_bytes;
let size = std::num::NonZeroU64::new(end - start).expect("buffer slice must not be empty");
self.buffer_bytes(buffer, start, Some(size))
}
/// Specify a sampler to be bound.
pub fn sampler(self, sampler: &'a wgpu::Sampler) -> Self {
let resource = wgpu::BindingResource::Sampler(sampler);
self.binding(resource)
}
/// Specify a texture view to be bound.
pub fn texture_view(self, view: &'a wgpu::TextureViewHandle) -> Self {
let resource = wgpu::BindingResource::TextureView(view);
self.binding(resource)
}
/// Build the bind group with the specified resources.
pub fn build(self, device: &wgpu::Device, layout: &wgpu::BindGroupLayout) -> wgpu::BindGroup {
let mut entries = Vec::with_capacity(self.resources.len());
for (i, resource) in self.resources.into_iter().enumerate() {
let binding = wgpu::BindGroupEntry {
binding: i as u32,
resource,
};
entries.push(binding);
}
let descriptor = wgpu::BindGroupDescriptor {
label: Some("nannou bind group"),
layout,
entries: &entries,
};
device.create_bind_group(&descriptor)
}
}