use std::cmp::Ordering; use crate::{ texture_atlas::{TextureAtlas, TextureAtlasSprite}, Sprite, SPRITE_SHADER_HANDLE, }; use bevy_asset::{AssetEvent, Assets, Handle, HandleId}; use bevy_core_pipeline::{core_2d::Transparent2d, tonemapping::Tonemapping}; use bevy_ecs::{ prelude::*, system::{lifetimeless::*, SystemParamItem, SystemState}, }; use bevy_math::{Rect, Vec2}; use bevy_reflect::Uuid; use bevy_render::{ color::Color, render_asset::RenderAssets, render_phase::{ BatchedPhaseItem, DrawFunctions, PhaseItem, RenderCommand, RenderCommandResult, RenderPhase, SetItemPipeline, TrackedRenderPass, }, render_resource::*, renderer::{RenderDevice, RenderQueue}, texture::{ BevyDefault, DefaultImageSampler, GpuImage, Image, ImageSampler, TextureFormatPixelInfo, }, view::{ ComputedVisibility, ExtractedView, Msaa, ViewTarget, ViewUniform, ViewUniformOffset, ViewUniforms, VisibleEntities, }, Extract, }; use bevy_transform::components::GlobalTransform; use bevy_utils::FloatOrd; use bevy_utils::HashMap; use bytemuck::{Pod, Zeroable}; use fixedbitset::FixedBitSet; #[derive(Resource)] pub struct SpritePipeline { view_layout: BindGroupLayout, material_layout: BindGroupLayout, pub dummy_white_gpu_image: GpuImage, } impl FromWorld for SpritePipeline { fn from_world(world: &mut World) -> Self { let mut system_state: SystemState<( Res, Res, Res, )> = SystemState::new(world); let (render_device, default_sampler, render_queue) = system_state.get_mut(world); let view_layout = render_device.create_bind_group_layout(&BindGroupLayoutDescriptor { entries: &[BindGroupLayoutEntry { binding: 0, visibility: ShaderStages::VERTEX | ShaderStages::FRAGMENT, ty: BindingType::Buffer { ty: BufferBindingType::Uniform, has_dynamic_offset: true, min_binding_size: Some(ViewUniform::min_size()), }, count: None, }], label: Some("sprite_view_layout"), }); let material_layout = render_device.create_bind_group_layout(&BindGroupLayoutDescriptor { entries: &[ BindGroupLayoutEntry { binding: 0, visibility: ShaderStages::FRAGMENT, ty: BindingType::Texture { multisampled: false, sample_type: TextureSampleType::Float { filterable: true }, view_dimension: TextureViewDimension::D2, }, count: None, }, BindGroupLayoutEntry { binding: 1, visibility: ShaderStages::FRAGMENT, ty: BindingType::Sampler(SamplerBindingType::Filtering), count: None, }, ], label: Some("sprite_material_layout"), }); let dummy_white_gpu_image = { let image = Image::new_fill( Extent3d::default(), TextureDimension::D2, &[255u8; 4], TextureFormat::bevy_default(), ); let texture = render_device.create_texture(&image.texture_descriptor); let sampler = match image.sampler_descriptor { ImageSampler::Default => (**default_sampler).clone(), ImageSampler::Descriptor(descriptor) => render_device.create_sampler(&descriptor), }; let format_size = image.texture_descriptor.format.pixel_size(); render_queue.write_texture( ImageCopyTexture { texture: &texture, mip_level: 0, origin: Origin3d::ZERO, aspect: TextureAspect::All, }, &image.data, ImageDataLayout { offset: 0, bytes_per_row: Some( std::num::NonZeroU32::new( image.texture_descriptor.size.width * format_size as u32, ) .unwrap(), ), rows_per_image: None, }, image.texture_descriptor.size, ); let texture_view = texture.create_view(&TextureViewDescriptor::default()); GpuImage { texture, texture_view, texture_format: image.texture_descriptor.format, sampler, size: Vec2::new( image.texture_descriptor.size.width as f32, image.texture_descriptor.size.height as f32, ), } }; SpritePipeline { view_layout, material_layout, dummy_white_gpu_image, } } } bitflags::bitflags! { #[repr(transparent)] // NOTE: Apparently quadro drivers support up to 64x MSAA. // MSAA uses the highest 3 bits for the MSAA log2(sample count) to support up to 128x MSAA. pub struct SpritePipelineKey: u32 { const NONE = 0; const COLORED = (1 << 0); const HDR = (1 << 1); const TONEMAP_IN_SHADER = (1 << 2); const DEBAND_DITHER = (1 << 3); const MSAA_RESERVED_BITS = Self::MSAA_MASK_BITS << Self::MSAA_SHIFT_BITS; } } impl SpritePipelineKey { const MSAA_MASK_BITS: u32 = 0b111; const MSAA_SHIFT_BITS: u32 = 32 - Self::MSAA_MASK_BITS.count_ones(); #[inline] pub const fn from_msaa_samples(msaa_samples: u32) -> Self { let msaa_bits = (msaa_samples.trailing_zeros() & Self::MSAA_MASK_BITS) << Self::MSAA_SHIFT_BITS; Self::from_bits_truncate(msaa_bits) } #[inline] pub const fn msaa_samples(&self) -> u32 { 1 << ((self.bits >> Self::MSAA_SHIFT_BITS) & Self::MSAA_MASK_BITS) } #[inline] pub const fn from_colored(colored: bool) -> Self { if colored { SpritePipelineKey::COLORED } else { SpritePipelineKey::NONE } } #[inline] pub const fn from_hdr(hdr: bool) -> Self { if hdr { SpritePipelineKey::HDR } else { SpritePipelineKey::NONE } } } impl SpecializedRenderPipeline for SpritePipeline { type Key = SpritePipelineKey; fn specialize(&self, key: Self::Key) -> RenderPipelineDescriptor { let mut formats = vec![ // position VertexFormat::Float32x3, // uv VertexFormat::Float32x2, ]; if key.contains(SpritePipelineKey::COLORED) { // color formats.push(VertexFormat::Float32x4); } let vertex_layout = VertexBufferLayout::from_vertex_formats(VertexStepMode::Vertex, formats); let mut shader_defs = Vec::new(); if key.contains(SpritePipelineKey::COLORED) { shader_defs.push("COLORED".into()); } if key.contains(SpritePipelineKey::TONEMAP_IN_SHADER) { shader_defs.push("TONEMAP_IN_SHADER".into()); // Debanding is tied to tonemapping in the shader, cannot run without it. if key.contains(SpritePipelineKey::DEBAND_DITHER) { shader_defs.push("DEBAND_DITHER".into()); } } let format = match key.contains(SpritePipelineKey::HDR) { true => ViewTarget::TEXTURE_FORMAT_HDR, false => TextureFormat::bevy_default(), }; RenderPipelineDescriptor { vertex: VertexState { shader: SPRITE_SHADER_HANDLE.typed::(), entry_point: "vertex".into(), shader_defs: shader_defs.clone(), buffers: vec![vertex_layout], }, fragment: Some(FragmentState { shader: SPRITE_SHADER_HANDLE.typed::(), shader_defs, entry_point: "fragment".into(), targets: vec![Some(ColorTargetState { format, blend: Some(BlendState::ALPHA_BLENDING), write_mask: ColorWrites::ALL, })], }), layout: Some(vec![self.view_layout.clone(), self.material_layout.clone()]), primitive: PrimitiveState { front_face: FrontFace::Ccw, cull_mode: None, unclipped_depth: false, polygon_mode: PolygonMode::Fill, conservative: false, topology: PrimitiveTopology::TriangleList, strip_index_format: None, }, depth_stencil: None, multisample: MultisampleState { count: key.msaa_samples(), mask: !0, alpha_to_coverage_enabled: false, }, label: Some("sprite_pipeline".into()), } } } #[derive(Component, Clone, Copy)] pub struct ExtractedSprite { pub entity: Entity, pub transform: GlobalTransform, pub color: Color, /// Select an area of the texture pub rect: Option, /// Change the on-screen size of the sprite pub custom_size: Option, /// Handle to the `Image` of this sprite /// PERF: storing a `HandleId` instead of `Handle` enables some optimizations (`ExtractedSprite` becomes `Copy` and doesn't need to be dropped) pub image_handle_id: HandleId, pub flip_x: bool, pub flip_y: bool, pub anchor: Vec2, } #[derive(Resource, Default)] pub struct ExtractedSprites { pub sprites: Vec, } #[derive(Resource, Default)] pub struct SpriteAssetEvents { pub images: Vec>, } pub fn extract_sprite_events( mut events: ResMut, mut image_events: Extract>>, ) { let SpriteAssetEvents { ref mut images } = *events; images.clear(); for image in image_events.iter() { // AssetEvent: !Clone images.push(match image { AssetEvent::Created { handle } => AssetEvent::Created { handle: handle.clone_weak(), }, AssetEvent::Modified { handle } => AssetEvent::Modified { handle: handle.clone_weak(), }, AssetEvent::Removed { handle } => AssetEvent::Removed { handle: handle.clone_weak(), }, }); } } pub fn extract_sprites( mut extracted_sprites: ResMut, texture_atlases: Extract>>, sprite_query: Extract< Query<( Entity, &ComputedVisibility, &Sprite, &GlobalTransform, &Handle, )>, >, atlas_query: Extract< Query<( Entity, &ComputedVisibility, &TextureAtlasSprite, &GlobalTransform, &Handle, )>, >, ) { extracted_sprites.sprites.clear(); for (entity, visibility, sprite, transform, handle) in sprite_query.iter() { if !visibility.is_visible() { continue; } // PERF: we don't check in this function that the `Image` asset is ready, since it should be in most cases and hashing the handle is expensive extracted_sprites.sprites.push(ExtractedSprite { entity, color: sprite.color, transform: *transform, rect: sprite.rect, // Pass the custom size custom_size: sprite.custom_size, flip_x: sprite.flip_x, flip_y: sprite.flip_y, image_handle_id: handle.id(), anchor: sprite.anchor.as_vec(), }); } for (entity, visibility, atlas_sprite, transform, texture_atlas_handle) in atlas_query.iter() { if !visibility.is_visible() { continue; } if let Some(texture_atlas) = texture_atlases.get(texture_atlas_handle) { let rect = Some(texture_atlas.textures[atlas_sprite.index]); extracted_sprites.sprites.push(ExtractedSprite { entity, color: atlas_sprite.color, transform: *transform, // Select the area in the texture atlas rect, // Pass the custom size custom_size: atlas_sprite.custom_size, flip_x: atlas_sprite.flip_x, flip_y: atlas_sprite.flip_y, image_handle_id: texture_atlas.texture.id(), anchor: atlas_sprite.anchor.as_vec(), }); } } } #[repr(C)] #[derive(Copy, Clone, Pod, Zeroable)] struct SpriteVertex { pub position: [f32; 3], pub uv: [f32; 2], } #[repr(C)] #[derive(Copy, Clone, Pod, Zeroable)] struct ColoredSpriteVertex { pub position: [f32; 3], pub uv: [f32; 2], pub color: [f32; 4], } #[derive(Resource)] pub struct SpriteMeta { vertices: BufferVec, colored_vertices: BufferVec, view_bind_group: Option, } impl Default for SpriteMeta { fn default() -> Self { Self { vertices: BufferVec::new(BufferUsages::VERTEX), colored_vertices: BufferVec::new(BufferUsages::VERTEX), view_bind_group: None, } } } const QUAD_INDICES: [usize; 6] = [0, 2, 3, 0, 1, 2]; const QUAD_VERTEX_POSITIONS: [Vec2; 4] = [ Vec2::new(-0.5, -0.5), Vec2::new(0.5, -0.5), Vec2::new(0.5, 0.5), Vec2::new(-0.5, 0.5), ]; const QUAD_UVS: [Vec2; 4] = [ Vec2::new(0., 1.), Vec2::new(1., 1.), Vec2::new(1., 0.), Vec2::new(0., 0.), ]; #[derive(Component, Eq, PartialEq, Copy, Clone)] pub struct SpriteBatch { image_handle_id: HandleId, colored: bool, } #[derive(Resource, Default)] pub struct ImageBindGroups { values: HashMap, BindGroup>, } #[allow(clippy::too_many_arguments)] pub fn queue_sprites( mut commands: Commands, mut view_entities: Local, draw_functions: Res>, render_device: Res, render_queue: Res, mut sprite_meta: ResMut, view_uniforms: Res, sprite_pipeline: Res, mut pipelines: ResMut>, pipeline_cache: Res, mut image_bind_groups: ResMut, gpu_images: Res>, msaa: Res, mut extracted_sprites: ResMut, mut views: Query<( &mut RenderPhase, &VisibleEntities, &ExtractedView, Option<&Tonemapping>, )>, events: Res, ) { // If an image has changed, the GpuImage has (probably) changed for event in &events.images { match event { AssetEvent::Created { .. } => None, AssetEvent::Modified { handle } | AssetEvent::Removed { handle } => { image_bind_groups.values.remove(handle) } }; } let msaa_key = SpritePipelineKey::from_msaa_samples(msaa.samples); if let Some(view_binding) = view_uniforms.uniforms.binding() { let sprite_meta = &mut sprite_meta; // Clear the vertex buffers sprite_meta.vertices.clear(); sprite_meta.colored_vertices.clear(); sprite_meta.view_bind_group = Some(render_device.create_bind_group(&BindGroupDescriptor { entries: &[BindGroupEntry { binding: 0, resource: view_binding, }], label: Some("sprite_view_bind_group"), layout: &sprite_pipeline.view_layout, })); let draw_sprite_function = draw_functions.read().id::(); // Vertex buffer indices let mut index = 0; let mut colored_index = 0; // FIXME: VisibleEntities is ignored let extracted_sprites = &mut extracted_sprites.sprites; // Sort sprites by z for correct transparency and then by handle to improve batching // NOTE: This can be done independent of views by reasonably assuming that all 2D views look along the negative-z axis in world space extracted_sprites.sort_unstable_by(|a, b| { match a .transform .translation() .z .partial_cmp(&b.transform.translation().z) { Some(Ordering::Equal) | None => a.image_handle_id.cmp(&b.image_handle_id), Some(other) => other, } }); let image_bind_groups = &mut *image_bind_groups; for (mut transparent_phase, visible_entities, view, tonemapping) in &mut views { let mut view_key = SpritePipelineKey::from_hdr(view.hdr) | msaa_key; if let Some(Tonemapping::Enabled { deband_dither }) = tonemapping { if !view.hdr { view_key |= SpritePipelineKey::TONEMAP_IN_SHADER; if *deband_dither { view_key |= SpritePipelineKey::DEBAND_DITHER; } } } let pipeline = pipelines.specialize( &pipeline_cache, &sprite_pipeline, view_key | SpritePipelineKey::from_colored(false), ); let colored_pipeline = pipelines.specialize( &pipeline_cache, &sprite_pipeline, view_key | SpritePipelineKey::from_colored(true), ); view_entities.clear(); view_entities.extend(visible_entities.entities.iter().map(|e| e.index() as usize)); transparent_phase.items.reserve(extracted_sprites.len()); // Impossible starting values that will be replaced on the first iteration let mut current_batch = SpriteBatch { image_handle_id: HandleId::Id(Uuid::nil(), u64::MAX), colored: false, }; let mut current_batch_entity = Entity::PLACEHOLDER; let mut current_image_size = Vec2::ZERO; // Add a phase item for each sprite, and detect when successive items can be batched. // Spawn an entity with a `SpriteBatch` component for each possible batch. // Compatible items share the same entity. // Batches are merged later (in `batch_phase_system()`), so that they can be interrupted // by any other phase item (and they can interrupt other items from batching). for extracted_sprite in extracted_sprites.iter() { if !view_entities.contains(extracted_sprite.entity.index() as usize) { continue; } let new_batch = SpriteBatch { image_handle_id: extracted_sprite.image_handle_id, colored: extracted_sprite.color != Color::WHITE, }; if new_batch != current_batch { // Set-up a new possible batch if let Some(gpu_image) = gpu_images.get(&Handle::weak(new_batch.image_handle_id)) { current_batch = new_batch; current_image_size = Vec2::new(gpu_image.size.x, gpu_image.size.y); current_batch_entity = commands.spawn(current_batch).id(); image_bind_groups .values .entry(Handle::weak(current_batch.image_handle_id)) .or_insert_with(|| { render_device.create_bind_group(&BindGroupDescriptor { entries: &[ BindGroupEntry { binding: 0, resource: BindingResource::TextureView( &gpu_image.texture_view, ), }, BindGroupEntry { binding: 1, resource: BindingResource::Sampler(&gpu_image.sampler), }, ], label: Some("sprite_material_bind_group"), layout: &sprite_pipeline.material_layout, }) }); } else { // Skip this item if the texture is not ready continue; } } // Calculate vertex data for this item let mut uvs = QUAD_UVS; if extracted_sprite.flip_x { uvs = [uvs[1], uvs[0], uvs[3], uvs[2]]; } if extracted_sprite.flip_y { uvs = [uvs[3], uvs[2], uvs[1], uvs[0]]; } // By default, the size of the quad is the size of the texture let mut quad_size = current_image_size; // If a rect is specified, adjust UVs and the size of the quad if let Some(rect) = extracted_sprite.rect { let rect_size = rect.size(); for uv in &mut uvs { *uv = (rect.min + *uv * rect_size) / current_image_size; } quad_size = rect_size; } // Override the size if a custom one is specified if let Some(custom_size) = extracted_sprite.custom_size { quad_size = custom_size; } // Apply size and global transform let positions = QUAD_VERTEX_POSITIONS.map(|quad_pos| { extracted_sprite .transform .transform_point( ((quad_pos - extracted_sprite.anchor) * quad_size).extend(0.), ) .into() }); // These items will be sorted by depth with other phase items let sort_key = FloatOrd(extracted_sprite.transform.translation().z); // Store the vertex data and add the item to the render phase if current_batch.colored { for i in QUAD_INDICES { sprite_meta.colored_vertices.push(ColoredSpriteVertex { position: positions[i], uv: uvs[i].into(), color: extracted_sprite.color.as_linear_rgba_f32(), }); } let item_start = colored_index; colored_index += QUAD_INDICES.len() as u32; let item_end = colored_index; transparent_phase.add(Transparent2d { draw_function: draw_sprite_function, pipeline: colored_pipeline, entity: current_batch_entity, sort_key, batch_range: Some(item_start..item_end), }); } else { for i in QUAD_INDICES { sprite_meta.vertices.push(SpriteVertex { position: positions[i], uv: uvs[i].into(), }); } let item_start = index; index += QUAD_INDICES.len() as u32; let item_end = index; transparent_phase.add(Transparent2d { draw_function: draw_sprite_function, pipeline, entity: current_batch_entity, sort_key, batch_range: Some(item_start..item_end), }); } } } sprite_meta .vertices .write_buffer(&render_device, &render_queue); sprite_meta .colored_vertices .write_buffer(&render_device, &render_queue); } } pub type DrawSprite = ( SetItemPipeline, SetSpriteViewBindGroup<0>, SetSpriteTextureBindGroup<1>, DrawSpriteBatch, ); pub struct SetSpriteViewBindGroup; impl RenderCommand

for SetSpriteViewBindGroup { type Param = SRes; type ViewWorldQuery = Read; type ItemWorldQuery = (); fn render<'w>( _item: &P, view_uniform: &'_ ViewUniformOffset, _entity: (), sprite_meta: SystemParamItem<'w, '_, Self::Param>, pass: &mut TrackedRenderPass<'w>, ) -> RenderCommandResult { pass.set_bind_group( I, sprite_meta.into_inner().view_bind_group.as_ref().unwrap(), &[view_uniform.offset], ); RenderCommandResult::Success } } pub struct SetSpriteTextureBindGroup; impl RenderCommand

for SetSpriteTextureBindGroup { type Param = SRes; type ViewWorldQuery = (); type ItemWorldQuery = Read; fn render<'w>( _item: &P, _view: (), sprite_batch: &'_ SpriteBatch, image_bind_groups: SystemParamItem<'w, '_, Self::Param>, pass: &mut TrackedRenderPass<'w>, ) -> RenderCommandResult { let image_bind_groups = image_bind_groups.into_inner(); pass.set_bind_group( I, image_bind_groups .values .get(&Handle::weak(sprite_batch.image_handle_id)) .unwrap(), &[], ); RenderCommandResult::Success } } pub struct DrawSpriteBatch; impl RenderCommand

for DrawSpriteBatch { type Param = SRes; type ViewWorldQuery = (); type ItemWorldQuery = Read; fn render<'w>( item: &P, _view: (), sprite_batch: &'_ SpriteBatch, sprite_meta: SystemParamItem<'w, '_, Self::Param>, pass: &mut TrackedRenderPass<'w>, ) -> RenderCommandResult { let sprite_meta = sprite_meta.into_inner(); if sprite_batch.colored { pass.set_vertex_buffer(0, sprite_meta.colored_vertices.buffer().unwrap().slice(..)); } else { pass.set_vertex_buffer(0, sprite_meta.vertices.buffer().unwrap().slice(..)); } pass.draw(item.batch_range().as_ref().unwrap().clone(), 0..1); RenderCommandResult::Success } }