Struct CallFaultControl 

pub struct CallFaultControl { /* private fields */ }

Arms and reads a CallFaults from any thread.

Implementations

impl CallFaultControl

pub fn fail_next(&self, call: Call, count: u32, failure: InjectedFailure)

Fail the next count calls of call with failure, then reach the kernel again. A count of zero clears call.

Panics

If failure is not a failing HRESULT, which only InjectedFailure::Hresult can express: this seam injects failure, never success.

pub fn fail_until_cleared(&self, call: Call, failure: InjectedFailure)

Fail every call of call with failure until clear.

Panics

As fail_next.

pub fn fail_next_completions(&self, count: u32, failure: InjectedFailure)

Rewrite the result code of the next count completions popped to failure, the identity and everything else in them untouched. A count of zero clears.

Panics

As fail_next.

pub fn stall_submits(&self, duration: Duration)

Stall submits for duration, from the first submit stalled: each returns without reaching the kernel, so what is queued stays pending, until duration has passed and submits reach the kernel and run it. A duration of zero clears.

pub fn clear(&self)

Stop failing and stalling: every call reaches the kernel again. What was recorded is kept.

pub fn failures(&self, call: Call) -> Vec<ThreadId>

The thread that made each failed call of call so far, in order.

A test that arms a failure on one thread and wants to show it was answered on another – a pool thread draining the ring – compares these with its own.

Trait Implementations

impl Clone for CallFaultControl

fn clone(&self) -> Self

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for CallFaultControl

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

Formats the value using the given formatter.