Struct PeakNs 

#[repr(transparent)]
pub struct PeakNs(pub u64);

Lifetime high-water mark, nanoseconds. The kernel updates these as a max-against-prior in update_stats_* / update_se / set_next_entity paths (kernel/sched/stats.c, kernel/sched/fair.c); the value at any procfs read is the largest single window the thread has accumulated since its birth. Group reduction takes max across contributors so the rendered cell surfaces the worst single window any thread experienced over its lifetime.

Cross-thread vs cross-snapshot semantics

The Max reduction over a bucket of threads produces the worst single window observed across DIFFERENT tasks — task A’s wait_max and task B’s wait_max measure two distinct scheduling histories, and the bucket-level max picks whichever task experienced the worst case. The result belongs to that one worst task, not to the bucket as a whole; downstream consumers should read the rendered cell as “this bucket contained at least one task that saw N ns of wait” rather than “all tasks in this bucket saw at most N ns of wait” (which is the same shape, but a much weaker statement).

In COMPARE mode the per-thread PeakNs delta between two snapshots is peak_after - peak_before — the kernel only ever raises the field, so the delta is non-negative and represents the AMOUNT BY WHICH THE LIFETIME HIGH-WATER LINE ROSE during the (capture-A, capture-B) interval, NOT the magnitude of the worst event in that interval. A new scheduling window inside the interval only moves the high-water line if its own magnitude exceeds every prior window the task had ever experienced; if every interval event was strictly smaller than peak_before, the delta is zero even though events did occur. The delta is therefore not itself a PeakNs in the same sense as the lifetime reading — it is a difference of high-water marks. The bucket reduction takes max over those deltas, surfacing the worst rise across contributors during the interval; this can dramatically under-report transient bad windows that happened earlier in any contributor’s lifetime.

Summing peaks across threads is a category error — does not implement Summable. Implements Maxable.

Examples: wait_max, sleep_max, block_max, exec_max, slice_max.

u64 backing vs kernel s64

Of the *_max schedstat fields, only exec_max is typed s64 in struct sched_statistics (include/linux/sched.h); wait_max, sleep_max, block_max, and slice_max are u64. The capture pipeline parses every dotted-ms.ns value via parsed_ns_from_dotted in crate::ctprof, which returns Err(ParseDottedNs::Negative) on negative dotted values; the parse_sched closure maps that to None, and the capture-site unwrap_or(0) then collapses None to zero before the wrapper is constructed. The u64 backing here is therefore safe even for exec_max because the parser path guarantees non-negative input — NOT because every kernel-side field promises non-negative. Any new writer that bypasses parsed_ns_from_dotted must replicate its non-negative guard.

Tuple Fields

0: u64

Trait Implementations

impl Clone for PeakNs

fn clone(&self) -> PeakNs

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for PeakNs

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

Formats the value using the given formatter.

impl Default for PeakNs

fn default() -> PeakNs

Returns the “default value” for a type.

impl<'de> Deserialize<'de> for PeakNs

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Display for PeakNs

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

Formats the value using the given formatter.

impl From<PeakNs> for u64

fn from(v: PeakNs) -> Self

Converts to this type from the input type.

impl From<u64> for PeakNs

fn from(v: u64) -> Self

Converts to this type from the input type.

impl Hash for PeakNs

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 Maxable for PeakNs

fn max_across(items: impl IntoIterator<Item = Self>) -> Option<Self>

impl Ord for PeakNs

fn cmp(&self, other: &PeakNs) -> Ordering

This method returns an Ordering between self and other.

fn max(self, other: Self) -> Self

where Self: Sized,

Compares and returns the maximum of two values.

fn min(self, other: Self) -> Self

where Self: Sized,

Compares and returns the minimum of two values.

fn clamp(self, min: Self, max: Self) -> Self

where Self: Sized,

Restrict a value to a certain interval.

impl PartialEq for PeakNs

fn eq(&self, other: &PeakNs) -> 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 PartialOrd for PeakNs

fn partial_cmp(&self, other: &PeakNs) -> Option<Ordering>

This method returns an ordering between self and other values if one exists.

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

Tests less than (for self and other) and is used by the < operator.

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

Tests less than or equal to (for self and other) and is used by the <= operator.

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

Tests greater than (for self and other) and is used by the > operator.

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

Tests greater than or equal to (for self and other) and is used by the >= operator.

impl Serialize for PeakNs

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.

impl Copy for PeakNs

impl Eq for PeakNs

impl StructuralPartialEq for PeakNs