###### (This is the documentation for SDL3, which is under heavy development and the API is changing! [SDL2](https://wiki.libsdl.org/SDL2/) is the current stable version!) # SDL_AddTimer Call a callback function at a future time. ## Header File Defined in [](https://github.com/libsdl-org/SDL/blob/main/include/SDL3/SDL_timer.h) ## Syntax ```c SDL_TimerID SDL_AddTimer(Uint32 interval, SDL_TimerCallback callback, void *param); ``` ## Function Parameters | | | | ---------------- | ------------------------------------------------------------------------------------------------- | | **interval** | the timer delay, in milliseconds, passed to `callback` | | **callback** | the [SDL_TimerCallback](SDL_TimerCallback) function to call when the specified `interval` elapses | | **param** | a pointer that is passed to `callback` | ## Return Value Returns a timer ID or 0 if an error occurs; call [SDL_GetError](SDL_GetError)() for more information. ## Remarks If you use this function, you must pass [`SDL_INIT_TIMER`](SDL_INIT_TIMER) to [SDL_Init](SDL_Init)(). The callback function is passed the current timer interval and the user supplied parameter from the [SDL_AddTimer](SDL_AddTimer)() call and should return the next timer interval. If the value returned from the callback is 0, the timer is canceled. The callback is run on a separate thread. Timers take into account the amount of time it took to execute the callback. For example, if the callback took 250 ms to execute and returned 1000 (ms), the timer would only wait another 750 ms before its next iteration. Timing may be inexact due to OS scheduling. Be sure to note the current time with [SDL_GetTicksNS](SDL_GetTicksNS)() or [SDL_GetPerformanceCounter](SDL_GetPerformanceCounter)() in case your callback needs to adjust for variances. ## Version This function is available since SDL 3.0.0. ## Code Examples ```c Uint32 my_callbackfunc_1(Uint32 interval, void *param) { SDL_Event event; SDL_UserEvent userevent; /* In this example, our callback pushes an SDL_EVENT_USER event into the queue, and causes our callback to be called again at the same interval: */ userevent.type = SDL_EVENT_USER; userevent.code = 0; userevent.data1 = NULL; userevent.data2 = NULL; event.type = SDL_EVENT_USER; event.user = userevent; SDL_PushEvent(&event); return interval; } int my_callback_param; /* Start the timer; the callback below will be executed after the delay */ Uint32 delay = ((33 + 5) / 10) * 10; /* To round it down to the nearest 10 ms */ SDL_TimerID my_timer_id = SDL_AddTimer(delay, my_callbackfunc_1, &my_callback_param); /* ... */ ``` Note that it is possible to avoid the multithreading problems with SDL timers by giving to `userevent.data1` the address of a function you want to be executed and to `userevent.data2` its params, and then deal with it in the event loop. ```c void my_function(void *); /* with the same code as before: */ Uint32 my_callbackfunc_2(Uint32 interval, void *param) { SDL_Event event; SDL_UserEvent userevent; /* In this example, our callback pushes a function into the queue, and causes our callback to be called again at the same interval: */ userevent.type = SDL_EVENT_USER; userevent.code = 0; userevent.data1 = &my_function; userevent.data2 = param; event.type = SDL_EVENT_USER; event.user = userevent; SDL_PushEvent(&event); return(interval); } /* Now the event loop */ SDL_Event event; while (SDL_PollEvent (&event)) { switch(event.type) { case SDL_EVENT_USER: { /* and now we can call the function we wanted to call in the timer but couldn't because of the multithreading problems */ void (*p) (void*) = event.user.data1; p(event.user.data2); break; } /* ... */ } } ``` ## See Also * [SDL_RemoveTimer](SDL_RemoveTimer) ---- [CategoryAPI](CategoryAPI), [CategoryAPIFunction](CategoryAPIFunction), [CategoryTimer](CategoryTimer)