mirror of
https://github.com/bevyengine/bevy
synced 2024-12-19 09:33:06 +00:00
5adf831b42
This patch adds the infrastructure necessary for Bevy to support *bindless resources*, by adding a new `#[bindless]` attribute to `AsBindGroup`. Classically, only a single texture (or sampler, or buffer) can be attached to each shader binding. This means that switching materials requires breaking a batch and issuing a new drawcall, even if the mesh is otherwise identical. This adds significant overhead not only in the driver but also in `wgpu`, as switching bind groups increases the amount of validation work that `wgpu` must do. *Bindless resources* are the typical solution to this problem. Instead of switching bindings between each texture, the renderer instead supplies a large *array* of all textures in the scene up front, and the material contains an index into that array. This pattern is repeated for buffers and samplers as well. The renderer now no longer needs to switch binding descriptor sets while drawing the scene. Unfortunately, as things currently stand, this approach won't quite work for Bevy. Two aspects of `wgpu` conspire to make this ideal approach unacceptably slow: 1. In the DX12 backend, all binding arrays (bindless resources) must have a constant size declared in the shader, and all textures in an array must be bound to actual textures. Changing the size requires a recompile. 2. Changing even one texture incurs revalidation of all textures, a process that takes time that's linear in the total size of the binding array. This means that declaring a large array of textures big enough to encompass the entire scene is presently unacceptably slow. For example, if you declare 4096 textures, then `wgpu` will have to revalidate all 4096 textures if even a single one changes. This process can take multiple frames. To work around this problem, this PR groups bindless resources into small *slabs* and maintains a free list for each. The size of each slab for the bindless arrays associated with a material is specified via the `#[bindless(N)]` attribute. For instance, consider the following declaration: ```rust #[derive(AsBindGroup)] #[bindless(16)] struct MyMaterial { #[buffer(0)] color: Vec4, #[texture(1)] #[sampler(2)] diffuse: Handle<Image>, } ``` The `#[bindless(N)]` attribute specifies that, if bindless arrays are supported on the current platform, each resource becomes a binding array of N instances of that resource. So, for `MyMaterial` above, the `color` attribute is exposed to the shader as `binding_array<vec4<f32>, 16>`, the `diffuse` texture is exposed to the shader as `binding_array<texture_2d<f32>, 16>`, and the `diffuse` sampler is exposed to the shader as `binding_array<sampler, 16>`. Inside the material's vertex and fragment shaders, the applicable index is available via the `material_bind_group_slot` field of the `Mesh` structure. So, for instance, you can access the current color like so: ```wgsl // `uniform` binding arrays are a non-sequitur, so `uniform` is automatically promoted // to `storage` in bindless mode. @group(2) @binding(0) var<storage> material_color: binding_array<Color, 4>; ... @fragment fn fragment(in: VertexOutput) -> @location(0) vec4<f32> { let color = material_color[mesh[in.instance_index].material_bind_group_slot]; ... } ``` Note that portable shader code can't guarantee that the current platform supports bindless textures. Indeed, bindless mode is only available in Vulkan and DX12. The `BINDLESS` shader definition is available for your use to determine whether you're on a bindless platform or not. Thus a portable version of the shader above would look like: ```wgsl #ifdef BINDLESS @group(2) @binding(0) var<storage> material_color: binding_array<Color, 4>; #else // BINDLESS @group(2) @binding(0) var<uniform> material_color: Color; #endif // BINDLESS ... @fragment fn fragment(in: VertexOutput) -> @location(0) vec4<f32> { #ifdef BINDLESS let color = material_color[mesh[in.instance_index].material_bind_group_slot]; #else // BINDLESS let color = material_color; #endif // BINDLESS ... } ``` Importantly, this PR *doesn't* update `StandardMaterial` to be bindless. So, for example, `scene_viewer` will currently not run any faster. I intend to update `StandardMaterial` to use bindless mode in a follow-up patch. A new example, `shaders/shader_material_bindless`, has been added to demonstrate how to use this new feature. Here's a Tracy profile of `submit_graph_commands` of this patch and an additional patch (not submitted yet) that makes `StandardMaterial` use bindless. Red is those patches; yellow is `main`. The scene was Bistro Exterior with a hack that forces all textures to opaque. You can see a 1.47x mean speedup. ![Screenshot 2024-11-12 161713](https://github.com/user-attachments/assets/4334b362-42c8-4d64-9cfb-6835f019b95c) ## Migration Guide * `RenderAssets::prepare_asset` now takes an `AssetId` parameter. * Bin keys now have Bevy-specific material bind group indices instead of `wgpu` material bind group IDs, as part of the bindless change. Use the new `MaterialBindGroupAllocator` to map from bind group index to bind group ID.
194 lines
6.2 KiB
Rust
194 lines
6.2 KiB
Rust
//! A shader that binds several textures onto one
|
|
//! `binding_array<texture<f32>>` shader binding slot and sample non-uniformly.
|
|
|
|
use bevy::{
|
|
ecs::system::{lifetimeless::SRes, SystemParamItem},
|
|
prelude::*,
|
|
reflect::TypePath,
|
|
render::{
|
|
render_asset::RenderAssets,
|
|
render_resource::{
|
|
binding_types::{sampler, texture_2d},
|
|
*,
|
|
},
|
|
renderer::RenderDevice,
|
|
texture::{FallbackImage, GpuImage},
|
|
RenderApp,
|
|
},
|
|
};
|
|
use std::{num::NonZero, process::exit};
|
|
|
|
/// This example uses a shader source file from the assets subdirectory
|
|
const SHADER_ASSET_PATH: &str = "shaders/texture_binding_array.wgsl";
|
|
|
|
fn main() {
|
|
let mut app = App::new();
|
|
app.add_plugins((
|
|
DefaultPlugins.set(ImagePlugin::default_nearest()),
|
|
GpuFeatureSupportChecker,
|
|
MaterialPlugin::<BindlessMaterial>::default(),
|
|
))
|
|
.add_systems(Startup, setup)
|
|
.run();
|
|
}
|
|
|
|
const MAX_TEXTURE_COUNT: usize = 16;
|
|
const TILE_ID: [usize; 16] = [
|
|
19, 23, 4, 33, 12, 69, 30, 48, 10, 65, 40, 47, 57, 41, 44, 46,
|
|
];
|
|
|
|
struct GpuFeatureSupportChecker;
|
|
|
|
impl Plugin for GpuFeatureSupportChecker {
|
|
fn build(&self, _app: &mut App) {}
|
|
|
|
fn finish(&self, app: &mut App) {
|
|
let Some(render_app) = app.get_sub_app_mut(RenderApp) else {
|
|
return;
|
|
};
|
|
|
|
let render_device = render_app.world().resource::<RenderDevice>();
|
|
|
|
// Check if the device support the required feature. If not, exit the example.
|
|
// In a real application, you should setup a fallback for the missing feature
|
|
if !render_device
|
|
.features()
|
|
.contains(WgpuFeatures::SAMPLED_TEXTURE_AND_STORAGE_BUFFER_ARRAY_NON_UNIFORM_INDEXING)
|
|
{
|
|
error!(
|
|
"Render device doesn't support feature \
|
|
SAMPLED_TEXTURE_AND_STORAGE_BUFFER_ARRAY_NON_UNIFORM_INDEXING, \
|
|
which is required for texture binding arrays"
|
|
);
|
|
exit(1);
|
|
}
|
|
}
|
|
}
|
|
|
|
fn setup(
|
|
mut commands: Commands,
|
|
mut meshes: ResMut<Assets<Mesh>>,
|
|
mut materials: ResMut<Assets<BindlessMaterial>>,
|
|
asset_server: Res<AssetServer>,
|
|
) {
|
|
commands.spawn((
|
|
Camera3d::default(),
|
|
Transform::from_xyz(2.0, 2.0, 2.0).looking_at(Vec3::new(0.0, 0.0, 0.0), Vec3::Y),
|
|
));
|
|
|
|
// load 16 textures
|
|
let textures: Vec<_> = TILE_ID
|
|
.iter()
|
|
.map(|id| asset_server.load(format!("textures/rpg/tiles/generic-rpg-tile{id:0>2}.png")))
|
|
.collect();
|
|
|
|
// a cube with multiple textures
|
|
commands.spawn((
|
|
Mesh3d(meshes.add(Cuboid::default())),
|
|
MeshMaterial3d(materials.add(BindlessMaterial { textures })),
|
|
));
|
|
}
|
|
|
|
#[derive(Asset, TypePath, Debug, Clone)]
|
|
struct BindlessMaterial {
|
|
textures: Vec<Handle<Image>>,
|
|
}
|
|
|
|
impl AsBindGroup for BindlessMaterial {
|
|
type Data = ();
|
|
|
|
type Param = (SRes<RenderAssets<GpuImage>>, SRes<FallbackImage>);
|
|
|
|
fn as_bind_group(
|
|
&self,
|
|
layout: &BindGroupLayout,
|
|
render_device: &RenderDevice,
|
|
(image_assets, fallback_image): &mut SystemParamItem<'_, '_, Self::Param>,
|
|
) -> Result<PreparedBindGroup<Self::Data>, AsBindGroupError> {
|
|
// retrieve the render resources from handles
|
|
let mut images = vec![];
|
|
for handle in self.textures.iter().take(MAX_TEXTURE_COUNT) {
|
|
match image_assets.get(handle) {
|
|
Some(image) => images.push(image),
|
|
None => return Err(AsBindGroupError::RetryNextUpdate),
|
|
}
|
|
}
|
|
|
|
let fallback_image = &fallback_image.d2;
|
|
|
|
let textures = vec![&fallback_image.texture_view; MAX_TEXTURE_COUNT];
|
|
|
|
// convert bevy's resource types to WGPU's references
|
|
let mut textures: Vec<_> = textures.into_iter().map(|texture| &**texture).collect();
|
|
|
|
// fill in up to the first `MAX_TEXTURE_COUNT` textures and samplers to the arrays
|
|
for (id, image) in images.into_iter().enumerate() {
|
|
textures[id] = &*image.texture_view;
|
|
}
|
|
|
|
let bind_group = render_device.create_bind_group(
|
|
"bindless_material_bind_group",
|
|
layout,
|
|
&BindGroupEntries::sequential((&textures[..], &fallback_image.sampler)),
|
|
);
|
|
|
|
Ok(PreparedBindGroup {
|
|
bindings: BindingResources(vec![]),
|
|
bind_group,
|
|
data: (),
|
|
})
|
|
}
|
|
|
|
fn unprepared_bind_group(
|
|
&self,
|
|
_layout: &BindGroupLayout,
|
|
_render_device: &RenderDevice,
|
|
_param: &mut SystemParamItem<'_, '_, Self::Param>,
|
|
) -> Result<UnpreparedBindGroup<Self::Data>, AsBindGroupError> {
|
|
// we implement as_bind_group directly because
|
|
panic!("bindless texture arrays can't be owned")
|
|
// or rather, they can be owned, but then you can't make a `&'a [&'a TextureView]` from a vec of them in get_binding().
|
|
}
|
|
|
|
fn bind_group_layout_entries(_: &RenderDevice) -> Vec<BindGroupLayoutEntry>
|
|
where
|
|
Self: Sized,
|
|
{
|
|
BindGroupLayoutEntries::with_indices(
|
|
// The layout entries will only be visible in the fragment stage
|
|
ShaderStages::FRAGMENT,
|
|
(
|
|
// Screen texture
|
|
//
|
|
// @group(2) @binding(0) var textures: binding_array<texture_2d<f32>>;
|
|
(
|
|
0,
|
|
texture_2d(TextureSampleType::Float { filterable: true })
|
|
.count(NonZero::<u32>::new(MAX_TEXTURE_COUNT as u32).unwrap()),
|
|
),
|
|
// Sampler
|
|
//
|
|
// @group(2) @binding(1) var nearest_sampler: sampler;
|
|
//
|
|
// Note: as with textures, multiple samplers can also be bound
|
|
// onto one binding slot:
|
|
//
|
|
// ```
|
|
// sampler(SamplerBindingType::Filtering)
|
|
// .count(NonZero::<u32>::new(MAX_TEXTURE_COUNT as u32).unwrap()),
|
|
// ```
|
|
//
|
|
// One may need to pay attention to the limit of sampler binding
|
|
// amount on some platforms.
|
|
(1, sampler(SamplerBindingType::Filtering)),
|
|
),
|
|
)
|
|
.to_vec()
|
|
}
|
|
}
|
|
|
|
impl Material for BindlessMaterial {
|
|
fn fragment_shader() -> ShaderRef {
|
|
SHADER_ASSET_PATH.into()
|
|
}
|
|
}
|