Class IJobParallelForDeferExtensions

Extension class for the IJobParallelForDefer job type providing custom overloads for scheduling and running.

Inheritance

IJobParallelForDeferExtensions

Namespace: Unity.Jobs

Assembly: Unity.Collections.dll

Syntax

public static class IJobParallelForDeferExtensions

Methods

EarlyJobInit<T>()

Gathers and caches reflection data for the internal job system's managed bindings. Unity is responsible for calling this method - don't call it yourself.

Declaration

public static void EarlyJobInit<T>() where T : struct, IJobParallelForDefer

Type Parameters

Name	Description
T

Remarks

When the Jobs package is included in the project, Unity generates code to call EarlyJobInit at startup. This allows Burst compiled code to schedule jobs because the reflection part of initialization, which is not compatible with burst compiler constraints, has already happened in EarlyJobInit.

Note: While the Jobs package code generator handles this automatically for all closed job types, you must register those with generic arguments (like IJobParallelForDefer<MyJobType<T>>) manually for each specialization with [[Unity.Jobs.RegisterGenericJobTypeAttribute]].

ScheduleByRef<T>(ref T, int*, int, JobHandle)

Schedule the job for execution on worker threads. forEachCount is a pointer to the number of iterations, when dependsOn has completed. This API is unsafe, it is recommended to use the NativeList based Schedule method instead.

Declaration

public static JobHandle ScheduleByRef<T>(this ref T jobData, int* forEachCount, int innerloopBatchCount, JobHandle dependsOn = default) where T : struct, IJobParallelForDefer

Parameters

Type	Name	Description
T	jobData	The job and data to schedule. In this variant, the jobData is passed by reference, which may be necessary for unusually large job structs.
int*	forEachCount	*forEachCount is used as the iteration count.
int	innerloopBatchCount	Granularity in which workstealing is performed. A value of 32, means the job queue will steal 32 iterations and then perform them in an efficient inner loop.
JobHandle	dependsOn	Dependencies are used to ensure that a job executes on workerthreads after the dependency has completed execution. Making sure that two jobs reading or writing to same data do not run in parallel.

Returns

Type	Description
JobHandle	JobHandle The handle identifying the scheduled job. Can be used as a dependency for a later job or ensure completion on the main thread.

Type Parameters

Name	Description
T

ScheduleByRef<T, U>(ref T, NativeList<U>, int, JobHandle)

Schedule the job for execution on worker threads. list.Length is used as the iteration count. Note that it is required to embed the list on the job struct as well.

Declaration

public static JobHandle ScheduleByRef<T, U>(this ref T jobData, NativeList<U> list, int innerloopBatchCount, JobHandle dependsOn = default) where T : struct, IJobParallelForDefer where U : unmanaged

Parameters

Type	Name	Description
T	jobData	The job and data to schedule. In this variant, the jobData is passed by reference, which may be necessary for unusually large job structs.
NativeList<U>	list	list.Length is used as the iteration count.
int	innerloopBatchCount	Granularity in which workstealing is performed. A value of 32, means the job queue will steal 32 iterations and then perform them in an efficient inner loop.
JobHandle	dependsOn	Dependencies are used to ensure that a job executes on workerthreads after the dependency has completed execution. Making sure that two jobs reading or writing to same data do not run in parallel.

Returns

Type	Description
JobHandle	JobHandle The handle identifying the scheduled job. Can be used as a dependency for a later job or ensure completion on the main thread.

Type Parameters

Name	Description
T	Job type
U	List element type

Schedule<T>(T, int*, int, JobHandle)

Declaration

public static JobHandle Schedule<T>(this T jobData, int* forEachCount, int innerloopBatchCount, JobHandle dependsOn = default) where T : struct, IJobParallelForDefer

Parameters

Type	Name	Description
T	jobData	The job and data to schedule.
int*	forEachCount	*forEachCount is used as the iteration count.
int	innerloopBatchCount	Granularity in which workstealing is performed. A value of 32, means the job queue will steal 32 iterations and then perform them in an efficient inner loop.
JobHandle	dependsOn	Dependencies are used to ensure that a job executes on workerthreads after the dependency has completed execution. Making sure that two jobs reading or writing to same data do not run in parallel.

Returns

Type	Description
JobHandle	JobHandle The handle identifying the scheduled job. Can be used as a dependency for a later job or ensure completion on the main thread.

Type Parameters

Name	Description
T	Job type

Schedule<T, U>(T, NativeList<U>, int, JobHandle)

Schedule the job for execution on worker threads. list.Length is used as the iteration count. Note that it is required to embed the list on the job struct as well.

Declaration

public static JobHandle Schedule<T, U>(this T jobData, NativeList<U> list, int innerloopBatchCount, JobHandle dependsOn = default) where T : struct, IJobParallelForDefer where U : unmanaged

Parameters

Type	Name	Description
T	jobData	The job and data to schedule.
NativeList<U>	list	list.Length is used as the iteration count.
int	innerloopBatchCount	Granularity in which workstealing is performed. A value of 32, means the job queue will steal 32 iterations and then perform them in an efficient inner loop.
JobHandle	dependsOn	Dependencies are used to ensure that a job executes on workerthreads after the dependency has completed execution. Making sure that two jobs reading or writing to same data do not run in parallel.

Returns

Type	Description
JobHandle	JobHandle The handle identifying the scheduled job. Can be used as a dependency for a later job or ensure completion on the main thread.

Type Parameters

Name	Description
T	Job type
U	List element type