Struct PendingActivation 

pub struct PendingActivation {
    pub id: ActivationId,
    pub proc: ProcessId,
    pub arg: ResumeArg,
    pub reconv_iter: u32,
    pub logical_order: (SchedTime, Region, u64),
}

The canonical, thread-count-independent ordering key entry (scheduler.md §6; ledger SC-014, SC-029, DT-005, DT-042).

Fixed vs. scheduler.md’s literal logical_order: (SimTime, Region, u64) 3-tuple (review fix, DT-042): DT-042 itself records that scheduler.md’s design “describes discarding [the delta/reconvergence-iteration count] once the delta loop converges, which conflicts with ARCHITECTURE §4’s 5-tuple key requiring it everywhere” — and per this workspace’s spec precedence (ARCHITECTURE.md §4 wins over docs/design/*.md), reconv_iter must be threaded through every PendingActivation. Without it, two activations of the same process in two different delta-loop reconvergence passes at the same (SchedTime, Region) compared equal under the old 3-tuple, collapsing a real ordering distinction DT-042/SC-013 require. (The tuple’s first field is now SchedTime, not vj_abi::SimTime — Task 1, SC-023/SC-033/SC-129: see SchedTime’s own doc comment.) id is new for the same review pass: determinism::DeterminismController ::on_read/on_write and ResolvedBatch::completed all key on ActivationId, but nothing previously let a PendingActivation/WorkItem report which ActivationId it is — id is that missing correlation.

Fields

id: ActivationId

proc: ProcessId

arg: ResumeArg

reconv_iter: u32

The delta/reconvergence-loop iteration count within (SchedTime, Region) — the canonical 5-tuple’s third field (ARCHITECTURE §4 item 3; DT-042).

logical_order: (SchedTime, Region, u64)

Trait Implementations

impl Clone for PendingActivation

fn clone(&self) -> Self

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Copy for PendingActivation

impl Debug for PendingActivation

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

Formats the value using the given formatter.