Struct PixelsHost 

pub struct PixelsHost { /* private fields */ }

Renders the requests made on its thread, for a frame loop whose thread cannot wait for the GPU: drive it every frame while it is waiting. Requests made while one exists go to it.

It draws with a renderer of its own rather than the frame’s: one-shot pixels must know whether their render dropped geometry (a scene busier than the reserved memory), and the frame’s renderer reports that for whichever of its frames it last measured.

Implementations

impl PixelsHost

pub fn new() -> PixelsHost

A host for this thread’s requests, with no GPU yet: requests wait for set_gpu.

pub fn waiting(&self) -> bool

pub fn needs_gpu(&self) -> bool

Whether requests wait for a renderer: the host creates one (on the web, asynchronously) and hands it over with set_gpu.

pub fn draws_on(&self, device: &Device) -> bool

Whether it renders on device; a host whose device was lost and replaced hands it a renderer on the new one.

pub fn set_gpu(&mut self, device: Device, queue: Queue, renderer: Renderer)

Render from now on with renderer on device.

pub fn refuse(&mut self, why: &str)

Answer every waiting request with why: the host could not create its renderer.

pub fn drive(&mut self) -> bool

Move the requests on without waiting: start the next one, or look whether the GPU has copied the current one out and answer it. One request at a time, so the renderer’s memory report is that request’s. Whether it answered one: what asked for it reads it on the next frame, so the caller draws one.

Trait Implementations

impl Default for PixelsHost

fn default() -> PixelsHost

Returns the “default value” for a type.

impl Drop for PixelsHost

fn drop(&mut self)

Executes the destructor for this type.

fn pin_drop(self: Pin<&mut Self>)

🔬This is a nightly-only experimental API. (pin_ergonomics)

Execute the destructor for this type, but different to Drop::drop, it requires self to be pinned.