Struct SchedIdentity 

pub struct SchedIdentity {
    pub pid_to_task: BTreeMap<Pid, TaskId>,
    pub tid_to_task: BTreeMap<Tid, TaskId>,
    pub pid_to_pgid: BTreeMap<Pid, Pid>,
    pub pid_to_sid: BTreeMap<Pid, Pid>,
    pub pgid_members: BTreeMap<Pid, Vec<TaskId>>,
    pub sid_members: BTreeMap<Pid, Vec<TaskId>>,
    pub parent_of: BTreeMap<TaskId, TaskId>,
    pub children_of: BTreeMap<TaskId, Vec<TaskId>>,
}

Identity maps for the scheduler: PID/TID routing, process groups, session membership, and parent/child relationships.

Lives behind the SCHED_IDENTITY lock, separate from GLOBAL_SCHED_STATE so that syscall lookups (getpid, getpgid, setpgid, setsid, etc.) never contend with fork/exit or block/wake paths.

Fields

pid_to_task: BTreeMap<Pid, TaskId>

Map userspace PID -> internal TaskId (process leader in current model).

tid_to_task: BTreeMap<Tid, TaskId>

Map userspace TID -> internal TaskId (fast thread lookup).

pid_to_pgid: BTreeMap<Pid, Pid>

Map PID -> process group id.

pid_to_sid: BTreeMap<Pid, Pid>

Map PID -> session id.

pgid_members: BTreeMap<Pid, Vec<TaskId>>

Group membership index: pgid -> task ids.

sid_members: BTreeMap<Pid, Vec<TaskId>>

Session membership index: sid -> task ids.

parent_of: BTreeMap<TaskId, TaskId>

Parent relationship: child -> parent

children_of: BTreeMap<TaskId, Vec<TaskId>>

Children list: parent -> children

Implementations

impl SchedIdentity

pub const fn new() -> Self

Creates a new empty identity registry.