Enum EncodeDecisionBlocked 

pub enum EncodeDecisionBlocked {
    NoDevice,
    NoDeviceEncodeKernel,
    DeviceNotEngaged,
}

Why a Stage-3 encode deployment decision could not be made from a real device measurement (#988, #1412). Each variant is a state in which the 100_000 rows/sec/GPU target was neither established NOR refuted on a device — the decision is blocked on hardware, not green-washed from a CPU proxy.

Variants

NoDevice

No CUDA device on this host: the exact encode could not be measured on a device at all (a CPU rate cannot substitute — that was the #1412 defect).

NoDeviceEncodeKernel

A device is present but there is no device-resident exact-encode kernel, so the FULL per-row encode cannot be measured on the device. (The resident normal-equations solve in crate::encode_throughput is only ONE component of the encode, not the encode; a component measurement cannot decide the encode surrogate question — #988.)

DeviceNotEngaged

A device is present and a measurement was attempted, but the device path did not engage (false routing) — refused rather than reported as a pass.

Trait Implementations

impl Clone for EncodeDecisionBlocked

fn clone(&self) -> EncodeDecisionBlocked

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Copy for EncodeDecisionBlocked

impl Debug for EncodeDecisionBlocked

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

Formats the value using the given formatter.

impl Eq for EncodeDecisionBlocked

impl PartialEq for EncodeDecisionBlocked

fn eq(&self, other: &EncodeDecisionBlocked) -> 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 StructuralPartialEq for EncodeDecisionBlocked