IUMSThreadProxy Structure

 

An abstraction for a thread of execution. If you want your scheduler to be granted user-mode schedulable (UMS) threads, set the value for the scheduler policy element SchedulerKind to UmsThreadDefault, and implement the IUMSScheduler interface. UMS threads are only supported on 64-bit operating systems with version Windows 7 and higher.

struct IUMSThreadProxy : public IThreadProxy;

Public Methods

NameDescription
IUMSThreadProxy::EnterCriticalRegion MethodCalled in order to enter a critical region. When inside a critical region, the scheduler will not observe asynchronous blocking operations that happen during the region. This means that the scheduler will not be reentered for page faults, thread suspensions, kernel asynchronous procedure calls (APCs), and so forth, for a UMS thread.
IUMSThreadProxy::EnterHyperCriticalRegion MethodCalled in order to enter a hyper-critical region. When inside a hyper-critical region, the scheduler will not observe any blocking operations that happen during the region. This means the scheduler will not be reentered for blocking function calls, lock acquisition attempts which block, page faults, thread suspensions, kernel asynchronous procedure calls (APCs), and so forth, for a UMS thread.
IUMSThreadProxy::ExitCriticalRegion MethodCalled in order to exit a critical region.
IUMSThreadProxy::ExitHyperCriticalRegion MethodCalled in order to exit a hyper-critical region.
IUMSThreadProxy::GetCriticalRegionType MethodReturns what kind of critical region the thread proxy is within. Because hyper-critical regions are a superset of critical regions, if code has entered a critical region and then a hyper-critical region, InsideHyperCriticalRegion will be returned.

IThreadProxy

IUMSThreadProxy

Header: concrtrm.h

Namespace: concurrency

Called in order to enter a critical region. When inside a critical region, the scheduler will not observe asynchronous blocking operations that happen during the region. This means that the scheduler will not be reentered for page faults, thread suspensions, kernel asynchronous procedure calls (APCs), and so forth, for a UMS thread.

virtual int EnterCriticalRegion() = 0;

Return Value

The new depth of critical region. Critical regions are reentrant.

Called in order to enter a hyper-critical region. When inside a hyper-critical region, the scheduler will not observe any blocking operations that happen during the region. This means the scheduler will not be reentered for blocking function calls, lock acquisition attempts which block, page faults, thread suspensions, kernel asynchronous procedure calls (APCs), and so forth, for a UMS thread.

virtual int EnterHyperCriticalRegion() = 0;

Return Value

The new depth of hyper-critical region. Hyper-critical regions are reentrant.

Remarks

The scheduler must be extraordinarily careful about what methods it calls and what locks it acquires in such regions. If code in such a region blocks on a lock that is held by something the scheduler is responsible for scheduling, deadlock may ensue.

Called in order to exit a critical region.

virtual int ExitCriticalRegion() = 0;

Return Value

The new depth of critical region. Critical regions are reentrant.

Called in order to exit a hyper-critical region.

virtual int ExitHyperCriticalRegion() = 0;

Return Value

The new depth of hyper-critical region. Hyper-critical regions are reentrant.

Returns what kind of critical region the thread proxy is within. Because hyper-critical regions are a superset of critical regions, if code has entered a critical region and then a hyper-critical region, InsideHyperCriticalRegion will be returned.

virtual CriticalRegionType GetCriticalRegionType() const = 0;

Return Value

The type of critical region the thread proxy is within.

concurrency Namespace
IUMSScheduler Structure

Show: