Source code

Revision control

Copy as Markdown

Other Tools

use core::ops::Range;
use crate::{link_to_wgpu_docs, link_to_wgpu_item, Extent3d, Origin3d};
#[cfg(any(feature = "serde", test))]
use serde::{Deserialize, Serialize};
#[cfg(doc)]
use crate::{BindingType, Features};
mod external_image;
mod external_texture;
mod format;
pub use external_image::*;
pub use external_texture::*;
pub use format::*;
/// Dimensionality of a texture.
///
/// Corresponds to [WebGPU `GPUTextureDimension`](
#[repr(C)]
#[derive(Copy, Clone, Debug, Hash, Eq, PartialEq)]
#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
pub enum TextureDimension {
/// 1D texture
#[cfg_attr(feature = "serde", serde(rename = "1d"))]
D1,
/// 2D texture
#[cfg_attr(feature = "serde", serde(rename = "2d"))]
D2,
/// 3D texture
#[cfg_attr(feature = "serde", serde(rename = "3d"))]
D3,
}
/// Order in which texture data is laid out in memory.
#[derive(Clone, Copy, Default, Debug, PartialEq, Eq, Hash)]
pub enum TextureDataOrder {
/// The texture is laid out densely in memory as:
///
/// ```text
/// Layer0Mip0 Layer0Mip1 Layer0Mip2
/// Layer1Mip0 Layer1Mip1 Layer1Mip2
/// Layer2Mip0 Layer2Mip1 Layer2Mip2
/// ````
///
/// This is the layout used by dds files.
#[default]
LayerMajor,
/// The texture is laid out densely in memory as:
///
/// ```text
/// Layer0Mip0 Layer1Mip0 Layer2Mip0
/// Layer0Mip1 Layer1Mip1 Layer2Mip1
/// Layer0Mip2 Layer1Mip2 Layer2Mip2
/// ```
///
/// This is the layout used by ktx and ktx2 files.
MipMajor,
}
/// Dimensions of a particular texture view.
///
/// Corresponds to [WebGPU `GPUTextureViewDimension`](
#[repr(C)]
#[derive(Copy, Clone, Debug, Default, Hash, Eq, PartialEq)]
#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
pub enum TextureViewDimension {
/// A one dimensional texture. `texture_1d` in WGSL and `texture1D` in GLSL.
#[cfg_attr(feature = "serde", serde(rename = "1d"))]
D1,
/// A two dimensional texture. `texture_2d` in WGSL and `texture2D` in GLSL.
#[cfg_attr(feature = "serde", serde(rename = "2d"))]
#[default]
D2,
/// A two dimensional array texture. `texture_2d_array` in WGSL and `texture2DArray` in GLSL.
#[cfg_attr(feature = "serde", serde(rename = "2d-array"))]
D2Array,
/// A cubemap texture. `texture_cube` in WGSL and `textureCube` in GLSL.
#[cfg_attr(feature = "serde", serde(rename = "cube"))]
Cube,
/// A cubemap array texture. `texture_cube_array` in WGSL and `textureCubeArray` in GLSL.
#[cfg_attr(feature = "serde", serde(rename = "cube-array"))]
CubeArray,
/// A three dimensional texture. `texture_3d` in WGSL and `texture3D` in GLSL.
#[cfg_attr(feature = "serde", serde(rename = "3d"))]
D3,
}
impl TextureViewDimension {
/// Get the texture dimension required of this texture view dimension.
#[must_use]
pub fn compatible_texture_dimension(self) -> TextureDimension {
match self {
Self::D1 => TextureDimension::D1,
Self::D2 | Self::D2Array | Self::Cube | Self::CubeArray => TextureDimension::D2,
Self::D3 => TextureDimension::D3,
}
}
}
/// Selects a subset of the data a [`Texture`] holds.
///
/// Used in [texture views](TextureViewDescriptor) and
/// [texture copy operations](TexelCopyTextureInfo).
///
/// Corresponds to [WebGPU `GPUTextureAspect`](
///
#[doc = link_to_wgpu_item!(struct Texture)]
#[repr(C)]
#[derive(Copy, Clone, Debug, Default, Hash, Eq, PartialEq)]
#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
#[cfg_attr(feature = "serde", serde(rename_all = "kebab-case"))]
pub enum TextureAspect {
/// Depth, Stencil, and Color.
#[default]
All,
/// Stencil.
StencilOnly,
/// Depth.
DepthOnly,
/// Plane 0.
Plane0,
/// Plane 1.
Plane1,
/// Plane 2.
Plane2,
}
impl TextureAspect {
/// Returns the texture aspect for a given plane.
#[must_use]
pub fn from_plane(plane: u32) -> Option<Self> {
Some(match plane {
0 => Self::Plane0,
1 => Self::Plane1,
2 => Self::Plane2,
_ => return None,
})
}
/// Returns the plane for a given texture aspect.
#[must_use]
pub fn to_plane(&self) -> Option<u32> {
match self {
TextureAspect::Plane0 => Some(0),
TextureAspect::Plane1 => Some(1),
TextureAspect::Plane2 => Some(2),
_ => None,
}
}
}
bitflags::bitflags! {
/// Different ways that you can use a texture.
///
/// The usages determine what kind of memory the texture is allocated from and what
/// actions the texture can partake in.
///
/// Corresponds to [WebGPU `GPUTextureUsageFlags`](
#[repr(transparent)]
#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
#[cfg_attr(feature = "serde", serde(transparent))]
#[derive(Debug, Copy, Clone, PartialEq, Eq, Hash)]
pub struct TextureUsages: u32 {
//
// ---- Start numbering at 1 << 0 ----
//
// WebGPU features:
//
/// Allows a texture to be the source in a [`CommandEncoder::copy_texture_to_buffer`] or
/// [`CommandEncoder::copy_texture_to_texture`] operation.
const COPY_SRC = 1 << 0;
/// Allows a texture to be the destination in a [`CommandEncoder::copy_buffer_to_texture`],
/// [`CommandEncoder::copy_texture_to_texture`], or [`Queue::write_texture`] operation.
const COPY_DST = 1 << 1;
/// Allows a texture to be a [`BindingType::Texture`] in a bind group.
const TEXTURE_BINDING = 1 << 2;
/// Allows a texture to be a [`BindingType::StorageTexture`] in a bind group.
const STORAGE_BINDING = 1 << 3;
/// Allows a texture to be an output attachment of a render pass.
///
/// Consider adding [`TextureUsages::TRANSIENT`] if the contents are not reused.
const RENDER_ATTACHMENT = 1 << 4;
//
// ---- Restart Numbering for Native Features ---
//
// Native Features:
//
/// Allows a texture to be used with image atomics. Requires [`Features::TEXTURE_ATOMIC`].
const STORAGE_ATOMIC = 1 << 16;
/// Specifies the contents of this texture will not be used in another pass to potentially reduce memory usage and bandwidth.
///
/// No-op on platforms on platforms that do not benefit from transient textures.
/// Generally mobile and Apple chips care about this.
///
/// Incompatible with ALL other usages except [`TextureUsages::RENDER_ATTACHMENT`] and requires it.
///
/// Requires [`StoreOp::Discard`].
const TRANSIENT = 1 << 17;
}
}
bitflags::bitflags! {
/// Similar to `TextureUsages`, but used only for `CommandEncoder::transition_resources`.
#[derive(Debug, Copy, Clone, PartialEq, Eq, Hash)]
#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
#[cfg_attr(feature = "serde", serde(transparent))]
pub struct TextureUses: u16 {
/// The texture is in unknown state.
const UNINITIALIZED = 1 << 0;
/// Ready to present image to the surface.
const PRESENT = 1 << 1;
/// The source of a hardware copy.
/// cbindgen:ignore
const COPY_SRC = 1 << 2;
/// The destination of a hardware copy.
/// cbindgen:ignore
const COPY_DST = 1 << 3;
/// Read-only sampled or fetched resource.
const RESOURCE = 1 << 4;
/// The color target of a renderpass.
const COLOR_TARGET = 1 << 5;
/// Read-only depth stencil usage.
const DEPTH_STENCIL_READ = 1 << 6;
/// Read-write depth stencil usage
const DEPTH_STENCIL_WRITE = 1 << 7;
/// Read-only storage texture usage. Corresponds to a UAV in d3d, so is exclusive, despite being read only.
/// cbindgen:ignore
const STORAGE_READ_ONLY = 1 << 8;
/// Write-only storage texture usage.
/// cbindgen:ignore
const STORAGE_WRITE_ONLY = 1 << 9;
/// Read-write storage texture usage.
/// cbindgen:ignore
const STORAGE_READ_WRITE = 1 << 10;
/// Image atomic enabled storage.
/// cbindgen:ignore
const STORAGE_ATOMIC = 1 << 11;
/// Transient texture that may not have any backing memory. Not a resource state stored in the trackers, only used for passing down usages to create_texture.
const TRANSIENT = 1 << 12;
/// The combination of states that a texture may be in _at the same time_.
/// cbindgen:ignore
const INCLUSIVE = Self::COPY_SRC.bits() | Self::RESOURCE.bits() | Self::DEPTH_STENCIL_READ.bits() | Self::STORAGE_READ_ONLY.bits();
/// The combination of states that a texture must exclusively be in.
/// cbindgen:ignore
const EXCLUSIVE = Self::COPY_DST.bits() | Self::COLOR_TARGET.bits() | Self::DEPTH_STENCIL_WRITE.bits() | Self::STORAGE_WRITE_ONLY.bits() | Self::STORAGE_READ_WRITE.bits() | Self::STORAGE_ATOMIC.bits() | Self::PRESENT.bits();
/// The combination of all usages that the are guaranteed to be be ordered by the hardware.
/// If a usage is ordered, then if the texture state doesn't change between draw calls, there
/// are no barriers needed for synchronization.
/// cbindgen:ignore
const ORDERED = Self::INCLUSIVE.bits() | Self::COLOR_TARGET.bits() | Self::DEPTH_STENCIL_WRITE.bits() | Self::STORAGE_READ_ONLY.bits();
/// Flag used by the wgpu-core texture tracker to say a texture is in different states for every sub-resource
const COMPLEX = 1 << 13;
/// Flag used by the wgpu-core texture tracker to say that the tracker does not know the state of the sub-resource.
/// This is different from UNINITIALIZED as that says the tracker does know, but the texture has not been initialized.
const UNKNOWN = 1 << 14;
}
}
/// A texture transition for use with `CommandEncoder::transition_resources`.
#[derive(Clone, Debug)]
#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
pub struct TextureTransition<T> {
/// The texture to transition.
pub texture: T,
/// An optional selector to transition only part of the texture.
///
/// If None, the entire texture will be transitioned.
pub selector: Option<TextureSelector>,
/// The new state to transition to.
pub state: TextureUses,
}
/// Specifies a particular set of subresources in a texture.
#[derive(Clone, Debug, PartialEq, Eq)]
#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
pub struct TextureSelector {
/// Range of mips to use.
pub mips: Range<u32>,
/// Range of layers to use.
pub layers: Range<u32>,
}
/// Specific type of a sample in a texture binding.
///
/// Corresponds to [WebGPU `GPUTextureSampleType`](
#[derive(Clone, Copy, Debug, Eq, PartialEq, Hash)]
#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
pub enum TextureSampleType {
/// Sampling returns floats.
///
/// Example WGSL syntax:
/// ```rust,ignore
/// @group(0) @binding(0)
/// var t: texture_2d<f32>;
/// ```
///
/// Example GLSL syntax:
/// ```cpp,ignore
/// layout(binding = 0)
/// uniform texture2D t;
/// ```
Float {
/// If this is `false`, the texture can't be sampled with
/// a filtering sampler.
///
/// Even if this is `true`, it's possible to sample with
/// a **non-filtering** sampler.
filterable: bool,
},
/// Sampling does the depth reference comparison.
///
/// This is also compatible with a non-filtering sampler.
///
/// Example WGSL syntax:
/// ```rust,ignore
/// @group(0) @binding(0)
/// var t: texture_depth_2d;
/// ```
///
/// Example GLSL syntax:
/// ```cpp,ignore
/// layout(binding = 0)
/// uniform texture2DShadow t;
/// ```
Depth,
/// Sampling returns signed integers.
///
/// Example WGSL syntax:
/// ```rust,ignore
/// @group(0) @binding(0)
/// var t: texture_2d<i32>;
/// ```
///
/// Example GLSL syntax:
/// ```cpp,ignore
/// layout(binding = 0)
/// uniform itexture2D t;
/// ```
Sint,
/// Sampling returns unsigned integers.
///
/// Example WGSL syntax:
/// ```rust,ignore
/// @group(0) @binding(0)
/// var t: texture_2d<u32>;
/// ```
///
/// Example GLSL syntax:
/// ```cpp,ignore
/// layout(binding = 0)
/// uniform utexture2D t;
/// ```
Uint,
}
impl Default for TextureSampleType {
fn default() -> Self {
Self::Float { filterable: true }
}
}
/// Specific type of a sample in a texture binding.
///
/// For use in [`BindingType::StorageTexture`].
///
/// Corresponds to [WebGPU `GPUStorageTextureAccess`](
#[derive(Clone, Copy, Debug, Eq, PartialEq, Hash)]
#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
#[cfg_attr(feature = "serde", serde(rename_all = "kebab-case"))]
pub enum StorageTextureAccess {
/// The texture can only be written in the shader and it:
/// - may or may not be annotated with `write` (WGSL).
/// - must be annotated with `writeonly` (GLSL).
///
/// Example WGSL syntax:
/// ```rust,ignore
/// @group(0) @binding(0)
/// var my_storage_image: texture_storage_2d<r32float, write>;
/// ```
///
/// Example GLSL syntax:
/// ```cpp,ignore
/// layout(set=0, binding=0, r32f) writeonly uniform image2D myStorageImage;
/// ```
WriteOnly,
/// The texture can only be read in the shader and it must be annotated with `read` (WGSL) or
/// `readonly` (GLSL).
///
/// [`Features::TEXTURE_ADAPTER_SPECIFIC_FORMAT_FEATURES`] must be enabled to use this access
/// mode. This is a native-only extension.
///
/// Example WGSL syntax:
/// ```rust,ignore
/// @group(0) @binding(0)
/// var my_storage_image: texture_storage_2d<r32float, read>;
/// ```
///
/// Example GLSL syntax:
/// ```cpp,ignore
/// layout(set=0, binding=0, r32f) readonly uniform image2D myStorageImage;
/// ```
ReadOnly,
/// The texture can be both read and written in the shader and must be annotated with
/// `read_write` in WGSL.
///
/// [`Features::TEXTURE_ADAPTER_SPECIFIC_FORMAT_FEATURES`] must be enabled to use this access
/// mode. This is a nonstandard, native-only extension.
///
/// Example WGSL syntax:
/// ```rust,ignore
/// @group(0) @binding(0)
/// var my_storage_image: texture_storage_2d<r32float, read_write>;
/// ```
///
/// Example GLSL syntax:
/// ```cpp,ignore
/// layout(set=0, binding=0, r32f) uniform image2D myStorageImage;
/// ```
ReadWrite,
/// The texture can be both read and written in the shader via atomics and must be annotated
/// with `read_write` in WGSL.
///
/// [`Features::TEXTURE_ADAPTER_SPECIFIC_FORMAT_FEATURES`] must be enabled to use this access
/// mode. This is a nonstandard, native-only extension.
///
/// Example WGSL syntax:
/// ```rust,ignore
/// @group(0) @binding(0)
/// var my_storage_image: texture_storage_2d<r32uint, atomic>;
/// ```
Atomic,
}
/// Describes a [`TextureView`].
///
/// For use with [`Texture::create_view()`].
///
/// Corresponds to [WebGPU `GPUTextureViewDescriptor`](
///
#[doc = link_to_wgpu_item!(struct TextureView)]
#[doc = link_to_wgpu_docs!(["`Texture::create_view()`"]: "struct.Texture.html#method.create_view")]
#[derive(Clone, Debug, Default, Eq, PartialEq)]
pub struct TextureViewDescriptor<L> {
/// Debug label of the texture view. This will show up in graphics debuggers for easy identification.
pub label: L,
/// Format of the texture view. Either must be the same as the texture format or in the list
/// of `view_formats` in the texture's descriptor.
pub format: Option<TextureFormat>,
/// The dimension of the texture view. For 1D textures, this must be `D1`. For 2D textures it must be one of
/// `D2`, `D2Array`, `Cube`, and `CubeArray`. For 3D textures it must be `D3`
pub dimension: Option<TextureViewDimension>,
/// The allowed usage(s) for the texture view. Must be a subset of the usage flags of the texture.
/// If not provided, defaults to the full set of usage flags of the texture.
pub usage: Option<TextureUsages>,
/// Aspect of the texture. Color textures must be [`TextureAspect::All`].
pub aspect: TextureAspect,
/// Base mip level.
pub base_mip_level: u32,
/// Mip level count.
/// If `Some(count)`, `base_mip_level + count` must be less or equal to underlying texture mip count.
/// If `None`, considered to include the rest of the mipmap levels, but at least 1 in total.
pub mip_level_count: Option<u32>,
/// Base array layer.
pub base_array_layer: u32,
/// Layer count.
/// If `Some(count)`, `base_array_layer + count` must be less or equal to the underlying array count.
/// If `None`, considered to include the rest of the array layers, but at least 1 in total.
pub array_layer_count: Option<u32>,
}
/// Describes a [`Texture`](../wgpu/struct.Texture.html).
///
/// Corresponds to [WebGPU `GPUTextureDescriptor`](
#[repr(C)]
#[derive(Clone, Debug, PartialEq, Eq, Hash)]
#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
pub struct TextureDescriptor<L, V> {
/// Debug label of the texture. This will show up in graphics debuggers for easy identification.
pub label: L,
/// Size of the texture. All components must be greater than zero. For a
/// regular 1D/2D texture, the unused sizes will be 1. For 2DArray textures,
/// Z is the number of 2D textures in that array.
pub size: Extent3d,
/// Mip count of texture. For a texture with no extra mips, this must be 1.
pub mip_level_count: u32,
/// Sample count of texture. If this is not 1, texture must have [`BindingType::Texture::multisampled`] set to true.
pub sample_count: u32,
/// Dimensions of the texture.
pub dimension: TextureDimension,
/// Format of the texture.
pub format: TextureFormat,
/// Allowed usages of the texture. If used in other ways, the operation will panic.
pub usage: TextureUsages,
/// Specifies what view formats will be allowed when calling `Texture::create_view` on this texture.
///
/// View formats of the same format as the texture are always allowed.
///
/// Note: currently, only the srgb-ness is allowed to change. (ex: `Rgba8Unorm` texture + `Rgba8UnormSrgb` view)
pub view_formats: V,
}
impl<L, V> TextureDescriptor<L, V> {
/// Takes a closure and maps the label of the texture descriptor into another.
#[must_use]
pub fn map_label<K>(&self, fun: impl FnOnce(&L) -> K) -> TextureDescriptor<K, V>
where
V: Clone,
{
TextureDescriptor {
label: fun(&self.label),
size: self.size,
mip_level_count: self.mip_level_count,
sample_count: self.sample_count,
dimension: self.dimension,
format: self.format,
usage: self.usage,
view_formats: self.view_formats.clone(),
}
}
/// Maps the label and view formats of the texture descriptor into another.
#[must_use]
pub fn map_label_and_view_formats<K, M>(
&self,
l_fun: impl FnOnce(&L) -> K,
v_fun: impl FnOnce(V) -> M,
) -> TextureDescriptor<K, M>
where
V: Clone,
{
TextureDescriptor {
label: l_fun(&self.label),
size: self.size,
mip_level_count: self.mip_level_count,
sample_count: self.sample_count,
dimension: self.dimension,
format: self.format,
usage: self.usage,
view_formats: v_fun(self.view_formats.clone()),
}
}
/// Calculates the extent at a given mip level.
///
/// If the given mip level is larger than possible, returns None.
///
/// Treats the depth as part of the mipmaps. If calculating
/// for a 2DArray texture, which does not mipmap depth, set depth to 1.
///
/// ```rust
/// # use wgpu_types as wgpu;
/// # type TextureDescriptor<'a> = wgpu::TextureDescriptor<(), &'a [wgpu::TextureFormat]>;
/// let desc = TextureDescriptor {
/// label: (),
/// size: wgpu::Extent3d { width: 100, height: 60, depth_or_array_layers: 1 },
/// mip_level_count: 7,
/// sample_count: 1,
/// dimension: wgpu::TextureDimension::D3,
/// format: wgpu::TextureFormat::Rgba8Sint,
/// usage: wgpu::TextureUsages::empty(),
/// view_formats: &[],
/// };
///
/// assert_eq!(desc.mip_level_size(0), Some(wgpu::Extent3d { width: 100, height: 60, depth_or_array_layers: 1 }));
/// assert_eq!(desc.mip_level_size(1), Some(wgpu::Extent3d { width: 50, height: 30, depth_or_array_layers: 1 }));
/// assert_eq!(desc.mip_level_size(2), Some(wgpu::Extent3d { width: 25, height: 15, depth_or_array_layers: 1 }));
/// assert_eq!(desc.mip_level_size(3), Some(wgpu::Extent3d { width: 12, height: 7, depth_or_array_layers: 1 }));
/// assert_eq!(desc.mip_level_size(4), Some(wgpu::Extent3d { width: 6, height: 3, depth_or_array_layers: 1 }));
/// assert_eq!(desc.mip_level_size(5), Some(wgpu::Extent3d { width: 3, height: 1, depth_or_array_layers: 1 }));
/// assert_eq!(desc.mip_level_size(6), Some(wgpu::Extent3d { width: 1, height: 1, depth_or_array_layers: 1 }));
/// assert_eq!(desc.mip_level_size(7), None);
/// ```
#[must_use]
pub fn mip_level_size(&self, level: u32) -> Option<Extent3d> {
if level >= self.mip_level_count {
return None;
}
Some(self.size.mip_level_size(level, self.dimension))
}
/// Computes the render extent of this texture.
///
/// This is a low-level helper exported for use by wgpu-core.
///
///
/// # Panics
///
/// If the mip level is out of range.
#[doc(hidden)]
#[must_use]
pub fn compute_render_extent(&self, mip_level: u32, plane: Option<u32>) -> Extent3d {
let Extent3d {
width,
height,
depth_or_array_layers: _,
} = self.mip_level_size(mip_level).expect("invalid mip level");
let (w_subsampling, h_subsampling) = self.format.subsampling_factors(plane);
let width = width / w_subsampling;
let height = height / h_subsampling;
Extent3d {
width,
height,
depth_or_array_layers: 1,
}
}
/// Returns the number of array layers.
///
#[must_use]
pub fn array_layer_count(&self) -> u32 {
match self.dimension {
TextureDimension::D1 | TextureDimension::D3 => 1,
TextureDimension::D2 => self.size.depth_or_array_layers,
}
}
}
/// Describes a `Sampler`.
///
/// For use with `Device::create_sampler`.
///
/// Corresponds to [WebGPU `GPUSamplerDescriptor`](
#[derive(Clone, Debug, PartialEq)]
#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
pub struct SamplerDescriptor<L> {
/// Debug label of the sampler. This will show up in graphics debuggers for easy identification.
pub label: L,
/// How to deal with out of bounds accesses in the u (i.e. x) direction
pub address_mode_u: AddressMode,
/// How to deal with out of bounds accesses in the v (i.e. y) direction
pub address_mode_v: AddressMode,
/// How to deal with out of bounds accesses in the w (i.e. z) direction
pub address_mode_w: AddressMode,
/// How to filter the texture when it needs to be magnified (made larger)
pub mag_filter: FilterMode,
/// How to filter the texture when it needs to be minified (made smaller)
pub min_filter: FilterMode,
/// How to filter between mip map levels
pub mipmap_filter: MipmapFilterMode,
/// Minimum level of detail (i.e. mip level) to use
pub lod_min_clamp: f32,
/// Maximum level of detail (i.e. mip level) to use
pub lod_max_clamp: f32,
/// If this is enabled, this is a comparison sampler using the given comparison function.
pub compare: Option<crate::CompareFunction>,
/// Must be at least 1. If this is not 1, all filter modes must be linear.
pub anisotropy_clamp: u16,
/// Border color to use when `address_mode` is [`AddressMode::ClampToBorder`]
pub border_color: Option<SamplerBorderColor>,
}
impl<L: Default> Default for SamplerDescriptor<L> {
fn default() -> Self {
Self {
label: Default::default(),
address_mode_u: Default::default(),
address_mode_v: Default::default(),
address_mode_w: Default::default(),
mag_filter: Default::default(),
min_filter: Default::default(),
mipmap_filter: Default::default(),
lod_min_clamp: 0.0,
lod_max_clamp: 32.0,
compare: None,
anisotropy_clamp: 1,
border_color: None,
}
}
}
/// How edges should be handled in texture addressing.
///
/// Corresponds to [WebGPU `GPUAddressMode`](
#[repr(C)]
#[derive(Copy, Clone, Debug, Default, Hash, Eq, PartialEq)]
#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
#[cfg_attr(feature = "serde", serde(rename_all = "kebab-case"))]
pub enum AddressMode {
/// Clamp the value to the edge of the texture
///
/// -0.25 -> 0.0
/// 1.25 -> 1.0
#[default]
ClampToEdge = 0,
/// Repeat the texture in a tiling fashion
///
/// -0.25 -> 0.75
/// 1.25 -> 0.25
Repeat = 1,
/// Repeat the texture, mirroring it every repeat
///
/// -0.25 -> 0.25
/// 1.25 -> 0.75
MirrorRepeat = 2,
/// Clamp the value to the border of the texture
/// Requires feature [`Features::ADDRESS_MODE_CLAMP_TO_BORDER`]
///
/// -0.25 -> border
/// 1.25 -> border
ClampToBorder = 3,
}
/// Texel mixing mode when sampling between texels.
///
/// Corresponds to [WebGPU `GPUFilterMode`](
#[repr(C)]
#[derive(Copy, Clone, Debug, Default, Hash, Eq, PartialEq)]
#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
#[cfg_attr(feature = "serde", serde(rename_all = "kebab-case"))]
pub enum FilterMode {
/// Nearest neighbor sampling.
///
/// This creates a pixelated effect.
#[default]
Nearest = 0,
/// Linear Interpolation
///
/// This makes textures smooth but blurry.
Linear = 1,
}
/// Texel mixing mode when sampling between texels.
///
/// Corresponds to [WebGPU `GPUMipmapFilterMode`](
#[repr(C)]
#[derive(Copy, Clone, Debug, Default, Hash, Eq, PartialEq)]
#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
#[cfg_attr(feature = "serde", serde(rename_all = "kebab-case"))]
pub enum MipmapFilterMode {
/// Nearest neighbor sampling.
///
/// Return the value of the texel nearest to the texture coordinates.
#[default]
Nearest = 0,
/// Linear Interpolation
///
/// Select two texels in each dimension and return a linear interpolation between their values.
Linear = 1,
}
/// Color variation to use when sampler addressing mode is [`AddressMode::ClampToBorder`]
#[repr(C)]
#[derive(Copy, Clone, Debug, Eq, PartialEq, Hash)]
#[cfg_attr(feature = "serde", derive(serde::Serialize, serde::Deserialize))]
pub enum SamplerBorderColor {
/// [0, 0, 0, 0]
TransparentBlack,
/// [0, 0, 0, 1]
OpaqueBlack,
/// [1, 1, 1, 1]
OpaqueWhite,
/// On the Metal backend, this is equivalent to `TransparentBlack` for
/// textures that have an alpha component, and equivalent to `OpaqueBlack`
/// for textures that do not have an alpha component. On other backends,
/// this is equivalent to `TransparentBlack`. Requires
/// [`Features::ADDRESS_MODE_CLAMP_TO_ZERO`]. Not supported on the web.
Zero,
}
/// Layout of a texture in a buffer's memory.
///
/// The bytes per row and rows per image can be hard to figure out so here are some examples:
///
/// | Resolution | Format | Bytes per block | Pixels per block | Bytes per row | Rows per image |
/// |------------|--------|-----------------|------------------|----------------------------------------|------------------------------|
/// | 256x256 | RGBA8 | 4 | 1 * 1 * 1 | 256 * 4 = Some(1024) | None |
/// | 32x16x8 | RGBA8 | 4 | 1 * 1 * 1 | 32 * 4 = 128 padded to 256 = Some(256) | None |
/// | 256x256 | BC3 | 16 | 4 * 4 * 1 | 16 * (256 / 4) = 1024 = Some(1024) | None |
/// | 64x64x8 | BC3 | 16 | 4 * 4 * 1 | 16 * (64 / 4) = 256 = Some(256) | 64 / 4 = 16 = Some(16) |
///
/// Corresponds to [WebGPU `GPUTexelCopyBufferLayout`](
#[repr(C)]
#[derive(Clone, Copy, Debug, Default)]
#[cfg_attr(feature = "serde", derive(serde::Serialize, serde::Deserialize))]
pub struct TexelCopyBufferLayout {
/// Offset into the buffer that is the start of the texture. Must be a multiple of texture block size.
/// For non-compressed textures, this is 1.
pub offset: crate::BufferAddress,
/// Bytes per "row" in an image.
///
/// A row is one row of pixels or of compressed blocks in the x direction.
///
/// This value is required if there are multiple rows (i.e. height or depth is more than one pixel or pixel block for compressed textures)
///
/// Must be a multiple of 256 for [`CommandEncoder::copy_buffer_to_texture`][CEcbtt]
/// and [`CommandEncoder::copy_texture_to_buffer`][CEcttb]. You must manually pad the
/// image such that this is a multiple of 256. It will not affect the image data.
///
/// [`Queue::write_texture`][Qwt] does not have this requirement.
///
/// Must be a multiple of the texture block size. For non-compressed textures, this is 1.
///
#[doc = link_to_wgpu_docs!(["CEcbtt"]: "struct.CommandEncoder.html#method.copy_buffer_to_texture")]
#[doc = link_to_wgpu_docs!(["CEcttb"]: "struct.CommandEncoder.html#method.copy_texture_to_buffer")]
#[doc = link_to_wgpu_docs!(["Qwt"]: "struct.Queue.html#method.write_texture")]
pub bytes_per_row: Option<u32>,
/// "Rows" that make up a single "image".
///
/// A row is one row of pixels or of compressed blocks in the x direction.
///
/// An image is one layer in the z direction of a 3D image or 2DArray texture.
///
/// The amount of rows per image may be larger than the actual amount of rows of data.
///
/// Required if there are multiple images (i.e. the depth is more than one).
pub rows_per_image: Option<u32>,
}
/// View of a buffer which can be used to copy to/from a texture.
///
/// Corresponds to [WebGPU `GPUTexelCopyBufferInfo`](
#[repr(C)]
#[derive(Copy, Clone, Debug)]
#[cfg_attr(feature = "serde", derive(serde::Serialize, serde::Deserialize))]
pub struct TexelCopyBufferInfo<B> {
/// The buffer to be copied to/from.
pub buffer: B,
/// The layout of the texture data in this buffer.
pub layout: TexelCopyBufferLayout,
}
/// View of a texture which can be used to copy to/from a buffer/texture.
///
/// Corresponds to [WebGPU `GPUTexelCopyTextureInfo`](
#[repr(C)]
#[derive(Copy, Clone, Debug)]
#[cfg_attr(feature = "serde", derive(serde::Serialize, serde::Deserialize))]
pub struct TexelCopyTextureInfo<T> {
/// The texture to be copied to/from.
pub texture: T,
/// The target mip level of the texture.
pub mip_level: u32,
/// The base texel of the texture in the selected `mip_level`. Together
/// with the `copy_size` argument to copy functions, defines the
/// sub-region of the texture to copy.
#[cfg_attr(feature = "serde", serde(default))]
pub origin: Origin3d,
/// The copy aspect.
#[cfg_attr(feature = "serde", serde(default))]
pub aspect: TextureAspect,
}
impl<T> TexelCopyTextureInfo<T> {
/// Adds color space and premultiplied alpha information to make this
/// descriptor tagged.
pub fn to_tagged(
self,
color_space: PredefinedColorSpace,
premultiplied_alpha: bool,
) -> CopyExternalImageDestInfo<T> {
CopyExternalImageDestInfo {
texture: self.texture,
mip_level: self.mip_level,
origin: self.origin,
aspect: self.aspect,
color_space,
premultiplied_alpha,
}
}
}
/// Subresource range within an image
#[repr(C)]
#[derive(Clone, Copy, Debug, Default, Eq, PartialEq)]
#[cfg_attr(feature = "serde", derive(serde::Serialize, serde::Deserialize))]
#[cfg_attr(feature = "serde", serde(rename_all = "camelCase"))]
pub struct ImageSubresourceRange {
/// Aspect of the texture. Color textures must be [`TextureAspect::All`][TAA].
///
#[doc = link_to_wgpu_docs!(["TAA"]: "enum.TextureAspect.html#variant.All")]
pub aspect: TextureAspect,
/// Base mip level.
pub base_mip_level: u32,
/// Mip level count.
/// If `Some(count)`, `base_mip_level + count` must be less or equal to underlying texture mip count.
/// If `None`, considered to include the rest of the mipmap levels, but at least 1 in total.
pub mip_level_count: Option<u32>,
/// Base array layer.
pub base_array_layer: u32,
/// Layer count.
/// If `Some(count)`, `base_array_layer + count` must be less or equal to the underlying array count.
/// If `None`, considered to include the rest of the array layers, but at least 1 in total.
pub array_layer_count: Option<u32>,
}
impl ImageSubresourceRange {
/// Returns if the given range represents a full resource, with a texture of the given
/// layer count and mip count.
///
/// ```rust
/// # use wgpu_types as wgpu;
///
/// let range_none = wgpu::ImageSubresourceRange {
/// aspect: wgpu::TextureAspect::All,
/// base_mip_level: 0,
/// mip_level_count: None,
/// base_array_layer: 0,
/// array_layer_count: None,
/// };
/// assert_eq!(range_none.is_full_resource(wgpu::TextureFormat::Stencil8, 5, 10), true);
///
/// let range_some = wgpu::ImageSubresourceRange {
/// aspect: wgpu::TextureAspect::All,
/// base_mip_level: 0,
/// mip_level_count: Some(5),
/// base_array_layer: 0,
/// array_layer_count: Some(10),
/// };
/// assert_eq!(range_some.is_full_resource(wgpu::TextureFormat::Stencil8, 5, 10), true);
///
/// let range_mixed = wgpu::ImageSubresourceRange {
/// aspect: wgpu::TextureAspect::StencilOnly,
/// base_mip_level: 0,
/// // Only partial resource
/// mip_level_count: Some(3),
/// base_array_layer: 0,
/// array_layer_count: None,
/// };
/// assert_eq!(range_mixed.is_full_resource(wgpu::TextureFormat::Stencil8, 5, 10), false);
/// ```
#[must_use]
pub fn is_full_resource(
&self,
format: TextureFormat,
mip_levels: u32,
array_layers: u32,
) -> bool {
// Mip level count and array layer count need to deal with both the None and Some(count) case.
let mip_level_count = self.mip_level_count.unwrap_or(mip_levels);
let array_layer_count = self.array_layer_count.unwrap_or(array_layers);
let aspect_eq = Some(format) == format.aspect_specific_format(self.aspect);
let base_mip_level_eq = self.base_mip_level == 0;
let mip_level_count_eq = mip_level_count == mip_levels;
let base_array_layer_eq = self.base_array_layer == 0;
let array_layer_count_eq = array_layer_count == array_layers;
aspect_eq
&& base_mip_level_eq
&& mip_level_count_eq
&& base_array_layer_eq
&& array_layer_count_eq
}
/// Returns the mip level range of a subresource range describes for a specific texture.
#[must_use]
pub fn mip_range(&self, mip_level_count: u32) -> Range<u32> {
self.base_mip_level..match self.mip_level_count {
Some(mip_level_count) => self.base_mip_level + mip_level_count,
None => mip_level_count,
}
}
/// Returns the layer range of a subresource range describes for a specific texture.
#[must_use]
pub fn layer_range(&self, array_layer_count: u32) -> Range<u32> {
self.base_array_layer..match self.array_layer_count {
Some(array_layer_count) => self.base_array_layer + array_layer_count,
None => array_layer_count,
}
}
}
#[cfg(test)]
mod tests {
use super::*;
use crate::Extent3d;
#[test]
fn test_physical_size() {
let format = TextureFormat::Bc1RgbaUnormSrgb; // 4x4 blocks
assert_eq!(
Extent3d {
width: 7,
height: 7,
depth_or_array_layers: 1
}
.physical_size(format),
Extent3d {
width: 8,
height: 8,
depth_or_array_layers: 1
}
);
// Doesn't change, already aligned
assert_eq!(
Extent3d {
width: 8,
height: 8,
depth_or_array_layers: 1
}
.physical_size(format),
Extent3d {
width: 8,
height: 8,
depth_or_array_layers: 1
}
);
let format = TextureFormat::Astc {
block: AstcBlock::B8x5,
channel: AstcChannel::Unorm,
}; // 8x5 blocks
assert_eq!(
Extent3d {
width: 7,
height: 7,
depth_or_array_layers: 1
}
.physical_size(format),
Extent3d {
width: 8,
height: 10,
depth_or_array_layers: 1
}
);
}
#[test]
fn test_max_mips() {
// 1D
assert_eq!(
Extent3d {
width: 240,
height: 1,
depth_or_array_layers: 1
}
.max_mips(TextureDimension::D1),
1
);
// 2D
assert_eq!(
Extent3d {
width: 1,
height: 1,
depth_or_array_layers: 1
}
.max_mips(TextureDimension::D2),
1
);
assert_eq!(
Extent3d {
width: 60,
height: 60,
depth_or_array_layers: 1
}
.max_mips(TextureDimension::D2),
6
);
assert_eq!(
Extent3d {
width: 240,
height: 1,
depth_or_array_layers: 1000
}
.max_mips(TextureDimension::D2),
8
);
// 3D
assert_eq!(
Extent3d {
width: 16,
height: 30,
depth_or_array_layers: 60
}
.max_mips(TextureDimension::D3),
6
);
}
}