Enum SpawnPlacement 

#[non_exhaustive]
pub enum SpawnPlacement {
    RunnerCgroup,
    Cgroup(Cow<'static, str>),
}

Placement target for Op::Spawn.

The previous taxonomy had two ops (SpawnWorkers and SpawnHost) representing the two placement choices; the unified Op::Spawn variant parameterises the placement so the framework has ONE spawn op with the placement as data. SpawnPlacement is #[non_exhaustive]; further placements are added here rather than as new Op variants.

#[non_exhaustive]

SpawnPlacement is #[non_exhaustive] — see [crate::non_exhaustive] for the cross-crate pattern-match and construction rules shared by every such type.

Variants (Non-exhaustive)

Non-exhaustive enums could have additional variants added in future. Therefore, when matching against variants of non-exhaustive enums, an extra wildcard arm must be added to account for any future variants.

RunnerCgroup

Spawn workers in the spawner’s own cgroup — the test runner’s cgroup, NOT any managed workload cgroup declared via CgroupDef or Op::AddCgroup. The handler issues ZERO cgroup ops; the workers inherit whatever cgroup the test runner sits in.

Inside a guest VM the runner’s cgroup is typically the root (cgid=1), so RunnerCgroup workers appear in snapshots under the root cgroup rather than under your workload’s named hierarchy.

WorkSpec::workers_pct is rejected for this placement — there’s no managed cgroup whose cpuset would supply the percentage denominator. Use an explicit .workers(N) count, or switch to Cgroup(name) against a cgroup whose cpuset gives workers_pct a denominator.

Why “RunnerCgroup”?

The previous shape used SpawnHost — “host” referred to the spawner’s own cgroup (analogous to the scheduler-observability “host tasks vs workload tasks” distinction in sched_ext schedulers, e.g. mitosis’s cell 0). RunnerCgroup names the placement target precisely (the test-runner process’s cgroup) without the host-vs-guest-machine ambiguity that “host” carried.

Cgroup(Cow<'static, str>)

Spawn workers and move them into the named managed cgroup. The cgroup must already exist when the spawn op applies — declared via CgroupDef in Step.setup, via Op::AddCgroup / Op::AddCgroupDef earlier in the same step, or on the persistent Backdrop.

Implementations

impl SpawnPlacement

pub fn cgroup(name: impl Into<Cow<'static, str>>) -> Self

Construct SpawnPlacement::Cgroup from any string-like input (&'static str, String, Cow<'static, str>). Mirrors the [impl Into<Cow<'static, str>>] convention used by every other cgroup-name constructor on Op (Op::add_cgroup, Op::spawn_workers, Op::move_all_tasks, …) so callers pass "name" not "name".into().

pub const fn runner_cgroup() -> Self

Construct SpawnPlacement::RunnerCgroup. Const so it composes inside const scenarios + builds.

Trait Implementations

impl Clone for SpawnPlacement

fn clone(&self) -> SpawnPlacement

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for SpawnPlacement

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

Formats the value using the given formatter.

impl Hash for SpawnPlacement

fn hash<__H: Hasher>(&self, state: &mut __H)

Feeds this value into the given Hasher.

fn hash_slice<H>(data: &[Self], state: &mut H)

where H: Hasher, Self: Sized,

Feeds a slice of this type into the given Hasher.

impl PartialEq for SpawnPlacement

fn eq(&self, other: &SpawnPlacement) -> 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 Eq for SpawnPlacement

impl StructuralPartialEq for SpawnPlacement