Struct AsyncComputeQueue 

pub struct AsyncComputeQueue {
    pub total_enqueued: usize,
    pub total_executed: usize,
    /* private fields */
}

Models an async compute queue: a FIFO of DispatchBatches that can be submitted independently of the graphics queue.

On a GPU this enables overlapping async compute with graphics work. Here it is a simple in-memory queue used to test scheduling logic.

Fields

total_enqueued: usize

Total number of batches ever enqueued (for profiling).

total_executed: usize

Total number of batches ever drained (executed).

Implementations

impl AsyncComputeQueue

pub fn new() -> Self

Create an empty async compute queue.

pub fn submit(&mut self, batch: DispatchBatch)

Enqueue a dispatch batch.

pub fn flush(&mut self) -> usize

Execute (drain) all pending batches and return the count executed.

In a real backend this would flush the command buffer to the GPU. Here it simply clears the queue and updates counters.

pub fn pending(&self) -> usize

Number of batches currently waiting in the queue.

pub fn is_idle(&self) -> bool

Returns true when there are no pending batches.

Trait Implementations

impl Default for AsyncComputeQueue

fn default() -> Self

Returns the “default value” for a type.