arsa  2.7
Classes | Macros | Functions
SDL_atomic.h File Reference
#include "SDL_stdinc.h"
#include "SDL_platform.h"
#include "begin_code.h"
#include "close_code.h"

Go to the source code of this file.

Classes

struct  SDL_atomic_t
 A type representing an atomic integer value. It is a struct so people don't accidentally use numeric operations on it. More...
 

Macros

#define SDL_CompilerBarrier()   { SDL_SpinLock _tmp = 0; SDL_AtomicLock(&_tmp); SDL_AtomicUnlock(&_tmp); }
 
#define SDL_MemoryBarrierRelease()   SDL_CompilerBarrier()
 
#define SDL_MemoryBarrierAcquire()   SDL_CompilerBarrier()
 
#define SDL_AtomicIncRef(a)   SDL_AtomicAdd(a, 1)
 Increment an atomic variable used as a reference count. More...
 
#define SDL_AtomicDecRef(a)   (SDL_AtomicAdd(a, -1) == 1)
 Decrement an atomic variable used as a reference count. More...
 

Functions

DECLSPEC void SDLCALL SDL_MemoryBarrierReleaseFunction (void)
 
DECLSPEC void SDLCALL SDL_MemoryBarrierAcquireFunction (void)
 
DECLSPEC SDL_bool SDLCALL SDL_AtomicCAS (SDL_atomic_t *a, int oldval, int newval)
 Set an atomic variable to a new value if it is currently an old value. More...
 
DECLSPEC int SDLCALL SDL_AtomicSet (SDL_atomic_t *a, int v)
 Set an atomic variable to a value. More...
 
DECLSPEC int SDLCALL SDL_AtomicGet (SDL_atomic_t *a)
 Get the value of an atomic variable. More...
 
DECLSPEC int SDLCALL SDL_AtomicAdd (SDL_atomic_t *a, int v)
 Add to an atomic variable. More...
 
DECLSPEC SDL_bool SDLCALL SDL_AtomicCASPtr (void **a, void *oldval, void *newval)
 Set a pointer to a new value if it is currently an old value. More...
 
DECLSPEC void *SDLCALL SDL_AtomicSetPtr (void **a, void *v)
 Set a pointer to a value atomically. More...
 
DECLSPEC void *SDLCALL SDL_AtomicGetPtr (void **a)
 Get the value of a pointer atomically. More...
 

SDL AtomicLock

The atomic locks are efficient spinlocks using CPU instructions, but are vulnerable to starvation and can spin forever if a thread holding a lock has been terminated. For this reason you should minimize the code executed inside an atomic lock and never do expensive things like API or system calls while holding them.

The atomic locks are not safe to lock recursively.

Porting Note: The spin lock functions and type are required and can not be emulated because they are used in the atomic emulation code.

typedef int SDL_SpinLock
 
DECLSPEC SDL_bool SDLCALL SDL_AtomicTryLock (SDL_SpinLock *lock)
 Try to lock a spin lock by setting it to a non-zero value. More...
 
DECLSPEC void SDLCALL SDL_AtomicLock (SDL_SpinLock *lock)
 Lock a spin lock by setting it to a non-zero value. More...
 
DECLSPEC void SDLCALL SDL_AtomicUnlock (SDL_SpinLock *lock)
 Unlock a spin lock by setting it to 0. Always returns immediately. More...
 

Detailed Description

Atomic operations.

IMPORTANT: If you are not an expert in concurrent lockless programming, you should only be using the atomic lock and reference counting functions in this file. In all other cases you should be protecting your data structures with full mutexes.

The list of "safe" functions to use are: SDL_AtomicLock() SDL_AtomicUnlock() SDL_AtomicIncRef() SDL_AtomicDecRef()

Seriously, here be dragons! ^^^^^^^^^^^^^^^^^^^^^^^^^^^

You can find out a little more about lockless programming and the subtle issues that can arise here: http://msdn.microsoft.com/en-us/library/ee418650%28v=vs.85%29.aspx

There's also lots of good information here: http://www.1024cores.net/home/lock-free-algorithms http://preshing.com/

These operations may or may not actually be implemented using processor specific atomic operations. When possible they are implemented as true processor specific atomic operations. When that is not possible the are implemented using locks that do use the available atomic operations.

All of the atomic operations that modify memory are full memory barriers.

Definition in file SDL_atomic.h.

Macro Definition Documentation

◆ SDL_AtomicDecRef

#define SDL_AtomicDecRef (   a)    (SDL_AtomicAdd(a, -1) == 1)

Decrement an atomic variable used as a reference count.

Returns
SDL_TRUE if the variable reached zero after decrementing, SDL_FALSE otherwise

Definition at line 262 of file SDL_atomic.h.

◆ SDL_AtomicIncRef

#define SDL_AtomicIncRef (   a)    SDL_AtomicAdd(a, 1)

Increment an atomic variable used as a reference count.

Definition at line 252 of file SDL_atomic.h.

◆ SDL_CompilerBarrier

#define SDL_CompilerBarrier ( )    { SDL_SpinLock _tmp = 0; SDL_AtomicLock(&_tmp); SDL_AtomicUnlock(&_tmp); }

The compiler barrier prevents the compiler from reordering reads and writes to globally visible variables across the call.

Definition at line 132 of file SDL_atomic.h.

◆ SDL_MemoryBarrierAcquire

#define SDL_MemoryBarrierAcquire ( )    SDL_CompilerBarrier()

Definition at line 208 of file SDL_atomic.h.

◆ SDL_MemoryBarrierRelease

#define SDL_MemoryBarrierRelease ( )    SDL_CompilerBarrier()

Definition at line 207 of file SDL_atomic.h.

Typedef Documentation

◆ SDL_SpinLock

typedef int SDL_SpinLock

Definition at line 89 of file SDL_atomic.h.

Function Documentation

◆ SDL_AtomicAdd()

DECLSPEC int SDLCALL SDL_AtomicAdd ( SDL_atomic_t a,
int  v 
)

Add to an atomic variable.

Returns
The previous value of the atomic variable.
Note
This same style can be used for any number operation

◆ SDL_AtomicCAS()

DECLSPEC SDL_bool SDLCALL SDL_AtomicCAS ( SDL_atomic_t a,
int  oldval,
int  newval 
)

Set an atomic variable to a new value if it is currently an old value.

Returns
SDL_TRUE if the atomic variable was set, SDL_FALSE otherwise.
Note
If you don't know what this function is for, you shouldn't use it!

◆ SDL_AtomicCASPtr()

DECLSPEC SDL_bool SDLCALL SDL_AtomicCASPtr ( void **  a,
void oldval,
void newval 
)

Set a pointer to a new value if it is currently an old value.

Returns
SDL_TRUE if the pointer was set, SDL_FALSE otherwise.
Note
If you don't know what this function is for, you shouldn't use it!

◆ SDL_AtomicGet()

DECLSPEC int SDLCALL SDL_AtomicGet ( SDL_atomic_t a)

Get the value of an atomic variable.

◆ SDL_AtomicGetPtr()

DECLSPEC void* SDLCALL SDL_AtomicGetPtr ( void **  a)

Get the value of a pointer atomically.

◆ SDL_AtomicLock()

DECLSPEC void SDLCALL SDL_AtomicLock ( SDL_SpinLock lock)

Lock a spin lock by setting it to a non-zero value.

Parameters
lockPoints to the lock.

◆ SDL_AtomicSet()

DECLSPEC int SDLCALL SDL_AtomicSet ( SDL_atomic_t a,
int  v 
)

Set an atomic variable to a value.

Returns
The previous value of the atomic variable.

◆ SDL_AtomicSetPtr()

DECLSPEC void* SDLCALL SDL_AtomicSetPtr ( void **  a,
void v 
)

Set a pointer to a value atomically.

Returns
The previous value of the pointer.

◆ SDL_AtomicTryLock()

DECLSPEC SDL_bool SDLCALL SDL_AtomicTryLock ( SDL_SpinLock lock)

Try to lock a spin lock by setting it to a non-zero value.

Parameters
lockPoints to the lock.
Returns
SDL_TRUE if the lock succeeded, SDL_FALSE if the lock is already held.

◆ SDL_AtomicUnlock()

DECLSPEC void SDLCALL SDL_AtomicUnlock ( SDL_SpinLock lock)

Unlock a spin lock by setting it to 0. Always returns immediately.

Parameters
lockPoints to the lock.

◆ SDL_MemoryBarrierAcquireFunction()

DECLSPEC void SDLCALL SDL_MemoryBarrierAcquireFunction ( void  )

◆ SDL_MemoryBarrierReleaseFunction()

DECLSPEC void SDLCALL SDL_MemoryBarrierReleaseFunction ( void  )

Memory barriers are designed to prevent reads and writes from being reordered by the compiler and being seen out of order on multi-core CPUs.

A typical pattern would be for thread A to write some data and a flag, and for thread B to read the flag and get the data. In this case you would insert a release barrier between writing the data and the flag, guaranteeing that the data write completes no later than the flag is written, and you would insert an acquire barrier between reading the flag and reading the data, to ensure that all the reads associated with the flag have completed.

In this pattern you should always see a release barrier paired with an acquire barrier and you should gate the data reads/writes with a single flag variable.

For more information on these semantics, take a look at the blog post: http://preshing.com/20120913/acquire-and-release-semantics