Struct ForegroundProbeTiming 

pub struct ForegroundProbeTiming {
    pub queued: Duration,
    pub lock_wait: Duration,
    pub walk: Duration,
}

Where the wall-clock time in one foreground probe actually went.

A single total for observe_foreground mixes three things that lead to opposite conclusions about the same measurement: half a second of queued is idle waiting behind unrelated work on tokio’s blocking pool, while half a second of walk is this probe genuinely burning CPU on the OS process-table walk. walk is the only field of the three that is unambiguously CPU spent by this probe.

Fields

queued: Duration

Call until the blocking closure began running: tokio’s blocking-pool queue. Waiting, not work – a PTY read parked on a pool thread can put a probe behind it without either of them burning a cycle.

lock_wait: Duration

Inside the closure, waiting for the PTY mutex, which live PTY I/O also holds.

walk: Duration

The OS process-table walk itself. The only one of the three that is unambiguously CPU this probe spent.

Trait Implementations

impl Clone for ForegroundProbeTiming

fn clone(&self) -> ForegroundProbeTiming

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Copy for ForegroundProbeTiming

impl Debug for ForegroundProbeTiming

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

Formats the value using the given formatter.

impl Eq for ForegroundProbeTiming

impl PartialEq for ForegroundProbeTiming

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

Equality operator ==.

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

Inequality operator !=.

impl StructuralPartialEq for ForegroundProbeTiming