Enum ConcurrencyLowering 

pub enum ConcurrencyLowering {
    Ucontext,
    Stackless,
}

How the C-family backend lowers a task’s suspend/resume. One point in the (one-dimensional) knob lattice; pick with ConcurrencyLowering::for_target.

Variants

Ucontext

Stackful coroutines via ucontext. Simple codegen, native speed, arbitrary nesting; needs ucontext + a full stack per task. Host / RISC-V / Tenstorrent.

Stackless

A stackless state machine. Universal (device / WASM / embedded), tiny frames, cheap yields; complex codegen + a small per-resume dispatch cost. CUDA / WASM.

Implementations

impl ConcurrencyLowering

pub fn for_target(target: CTarget) -> Self

The best-general lowering each target CAN express. A target without ucontext is FORCED to ConcurrencyLowering::Stackless; a target that has it defaults to ConcurrencyLowering::Ucontext (simpler codegen + native speed — the right general call when portability is not at stake).

pub fn tradeoffs() -> &'static [ConcurrencyTradeoff]

The trade-off table — one row per setting, each naming where it wins and where it loses. The concurrency-knob lock asserts this stays complete and honest (neither setting is strictly better, which is exactly why this is a knob and not a baked-in choice).

Trait Implementations

impl Clone for ConcurrencyLowering

fn clone(&self) -> ConcurrencyLowering

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for ConcurrencyLowering

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl PartialEq for ConcurrencyLowering

fn eq(&self, other: &ConcurrencyLowering) -> bool

Tests for self and other values to be equal, and is used by ==.

fn ne(&self, other: &Rhs) -> bool

Tests for !=. The default implementation is almost always sufficient, and should not be overridden without very good reason.

impl Copy for ConcurrencyLowering

impl Eq for ConcurrencyLowering

impl StructuralPartialEq for ConcurrencyLowering