glibmm 2.66.5
Classes | Public Member Functions | Static Public Member Functions | Related Functions | List of all members
Glib::Thread Class Reference

Represents a running thread. More...

#include <glibmm/thread.h>

Classes

class  Exit
 Exception class used to exit from a thread. More...
 

Public Member Functions

 Thread (const Thread &)=delete
 
Threadoperator= (const Thread &)=delete
 
void join ()
 Waits until the thread finishes. More...
 
bool joinable () const
 Returns whether the thread is joinable. More...
 
void set_priority (ThreadPriority priority)
 Changes the priority of the thread to priority. More...
 
ThreadPriority get_priority () const
 Returns the priority of the thread. More...
 
GThread * gobj ()
 
const GThread * gobj () const
 

Static Public Member Functions

static Threadcreate (const sigc::slot< void > & slot, bool joinable=true)
 Creates a new thread with the priority THREAD_PRIORITY_NORMAL. More...
 
static Threadself ()
 Returns the Thread* corresponding to the calling thread. More...
 
static Threadcreate (const sigc::slot< void > & slot, unsigned long stack_size, bool joinable, bool bound, ThreadPriority priority)
 Creates a new thread with the priority priority. More...
 
static void yield ()
 Gives way to other threads waiting to be scheduled. More...
 

Related Functions

(Note that these are not member functions.)

Threadwrap (GThread * gobject)
 

Detailed Description

Represents a running thread.

An instance of this class can only be obtained with create(), self(), or wrap(GThread*). It's not possible to delete a Thread object. If the thread is not joinable, its resources will be freed automatically when it exits. Otherwise, if the thread is joinable, you must call join() to avoid a memory leak.

Note
g_thread_exit() is not wrapped, because that function exits a thread without any cleanup. That's especially dangerous in C++ code, since the destructors of automatic objects won't be invoked. Instead, you can throw a Thread::Exit exception, which will be caught by the internal thread entry function.
You might have noticed that the thread entry slot doesn't have the usual void* return value. If you want to return any data from your thread you can pass an additional output argument to the thread's entry slot.
Deprecated:
Use Glib::Threads::Thread instead.

Constructor & Destructor Documentation

◆ Thread()

Glib::Thread::Thread ( const Thread )
delete

Member Function Documentation

◆ create() [1/2]

static Thread * Glib::Thread::create ( const sigc::slot< void > &  slot,
bool  joinable = true 
)
static

Creates a new thread with the priority THREAD_PRIORITY_NORMAL.

If joinable is true, you can wait for this thread's termination by calling join(). Otherwise the thread will just disappear, when ready.

The new thread executes the function or method slot points to. You can pass additional arguments using sigc::bind(). If the thread was created successfully, it is returned, otherwise a ThreadError exception is thrown.

Because sigc::trackable is not thread safe, if the slot represents a non-static class method (that is, it is created by sigc::mem_fun()), the class concerned should not derive from sigc::trackable.

Parameters
slotA slot to execute in the new thread.
joinableThis parameter is now ignored because Threads are now always joinable.
Returns
The new Thread* on success.
Exceptions
Glib::ThreadError

◆ create() [2/2]

static Thread * Glib::Thread::create ( const sigc::slot< void > &  slot,
unsigned long  stack_size,
bool  joinable,
bool  bound,
ThreadPriority  priority 
)
static

Creates a new thread with the priority priority.

The stack gets the size stack_size or the default value for the current platform, if stack_size is 0.

If joinable is true, you can wait for this thread's termination by calling join(). Otherwise the thread will just disappear, when ready. If bound is true, this thread will be scheduled in the system scope, otherwise the implementation is free to do scheduling in the process scope. The first variant is more expensive resource-wise, but generally faster. On some systems (e.g. Linux) all threads are bound.

The new thread executes the function or method slot points to. You can pass additional arguments using sigc::bind(). If the thread was created successfully, it is returned.

Because sigc::trackable is not thread safe, if the slot represents a non-static class method (that is, it is created by sigc::mem_fun()), the class concerned should not derive from sigc::trackable.

Note
It is not guaranteed, that threads with different priorities really behave accordingly. On some systems (e.g. Linux) only root can increase priorities. On other systems (e.g. Solaris) there doesn't seem to be different scheduling for different priorities. All in all try to avoid being dependent on priorities. Use Glib::THREAD_PRIORITY_NORMAL here as a default.
Only use the extended create(const sigc::slot<void>&, unsigned long, bool, bool, ThreadPriority) function, when you really can't use the simple create(const sigc::slot<void>&, bool) instead. The latter overload does not take stack_size, bound and priority as arguments, as they should only be used for cases, where it is inevitable.
Parameters
slotA slot to execute in the new thread.
stack_sizeA stack size for the new thread, or 0.
joinableShould this thread be joinable?
boundShould this thread be bound to a system thread?
priorityA priority for the thread.
Returns
The new Thread* on success.
Exceptions
Glib::ThreadError
Deprecated:
Use the simpler create() method instead, because all Threads are now joinable, and bounds and priority parameters now have no effect.

◆ get_priority()

ThreadPriority Glib::Thread::get_priority ( ) const

Returns the priority of the thread.

Returns
The thread's priority.
Deprecated:
Thread priorities no longer have any effect.

◆ gobj() [1/2]

GThread * Glib::Thread::gobj ( )
inline

◆ gobj() [2/2]

const GThread * Glib::Thread::gobj ( ) const
inline

◆ join()

void Glib::Thread::join ( )

Waits until the thread finishes.

Waits until the thread finishes, i.e. the slot, as given to create(), returns or g_thread_exit() is called by the thread. (Calling g_thread_exit() in a C++ program should be avoided.) All resources of the thread including the Glib::Thread object are released. The thread must have been created with joinable = true.

◆ joinable()

bool Glib::Thread::joinable ( ) const

Returns whether the thread is joinable.

Returns
Whether the thread is joinable.
Deprecated:
All threads are now joinable.

◆ operator=()

Thread & Glib::Thread::operator= ( const Thread )
delete

◆ self()

static Thread * Glib::Thread::self ( )
static

Returns the Thread* corresponding to the calling thread.

Returns
The current thread.

◆ set_priority()

void Glib::Thread::set_priority ( ThreadPriority  priority)

Changes the priority of the thread to priority.

Note
It is not guaranteed, that threads with different priorities really behave accordingly. On some systems (e.g. Linux) only root can increase priorities. On other systems (e.g. Solaris) there doesn't seem to be different scheduling for different priorities. All in all try to avoid being dependent on priorities.
Parameters
priorityA new priority for the thread.
Deprecated:
Thread priorities no longer have any effect.

◆ yield()

static void Glib::Thread::yield ( )
static

Gives way to other threads waiting to be scheduled.

This function is often used as a method to make busy wait less evil. But in most cases, you will encounter, there are better methods to do that. So in general you shouldn't use this function.

Friends And Related Function Documentation

◆ wrap()

Thread * wrap ( GThread *  gobject)
related
Deprecated:
Use Glib::Threads::wrap(GThread*) instead.