std：：线程参数在转换为rvalues后必须是可调用的

Question

我正在尝试创建2个线程，它们与浮点数组一起工作。Howere编译器给了我这些错误：

/usr/include/c++/9/thread: In instantiation of ‘std::thread::thread(_Callable&&, _Args&& ...) [with _Callable = void (&)(bool*, float*, float*, float*); _Args = {bool*, float (*)[10], float (*)[10], float (*)[10]}; <template-parameter-1-3> = void]’:
main.cpp:16:66:   required from here
/usr/include/c++/9/thread:120:44: error: static assertion failed: std::thread arguments must be invocable after conversion to rvalues
  120 |           typename decay<_Args>::type...>::value,
      |                                            ^~~~~
/usr/include/c++/9/thread: In instantiation of ‘struct std::thread::_Invoker<std::tuple<void (*)(bool*, float*, float*, float*), bool*, float (*)[10], float (*)[10], float (*)[10]> >’:
/usr/include/c++/9/thread:131:22:   required from ‘std::thread::thread(_Callable&&, _Args&& ...) [with _Callable = void (&)(bool*, float*, float*, float*); _Args = {bool*, float (*)[10], float (*)[10], float (*)[10]}; <template-parameter-1-3> = void]’
main.cpp:16:66:   required from here
/usr/include/c++/9/thread:243:4: error: no type named ‘type’ in ‘struct std::thread::_Invoker<std::tuple<void (*)(bool*, float*, float*, float*), bool*, float (*)[10], float (*)[10], float (*)[10]> >::__result<std::tuple<void (*)(bool*, float*, float*, float*), bool*, float (*)[10], float (*)[10], float (*)[10]> >’
  243 |    _M_invoke(_Index_tuple<_Ind...>)
      |    ^~~~~~~~~
/usr/include/c++/9/thread:247:2: error: no type named ‘type’ in ‘struct std::thread::_Invoker<std::tuple<void (*)(bool*, float*, float*, float*), bool*, float (*)[10], float (*)[10], float (*)[10]> >::__result<std::tuple<void (*)(bool*, float*, float*, float*), bool*, float (*)[10], float (*)[10], float (*)[10]> >’
  247 |  operator()()
      |  ^~~~~~~~
/usr/include/c++/9/thread: In instantiation of ‘std::thread::thread(_Callable&&, _Args&& ...) [with _Callable = void (&)(float*); _Args = {float (*)[10]}; <template-parameter-1-3> = void]’:
main.cpp:17:43:   required from here
/usr/include/c++/9/thread:120:44: error: static assertion failed: std::thread arguments must be invocable after conversion to rvalues
  120 |           typename decay<_Args>::type...>::value,
      |                                            ^~~~~
/usr/include/c++/9/thread: In instantiation of ‘struct std::thread::_Invoker<std::tuple<void (*)(float*), float (*)[10]> >’:
/usr/include/c++/9/thread:131:22:   required from ‘std::thread::thread(_Callable&&, _Args&& ...) [with _Callable = void (&)(float*); _Args = {float (*)[10]}; <template-parameter-1-3> = void]’
main.cpp:17:43:   required from here
/usr/include/c++/9/thread:243:4: error: no type named ‘type’ in ‘struct std::thread::_Invoker<std::tuple<void (*)(float*), float (*)[10]> >::__result<std::tuple<void (*)(float*), float (*)[10]> >’
  243 |    _M_invoke(_Index_tuple<_Ind...>)
      |    ^~~~~~~~~
/usr/include/c++/9/thread:247:2: error: no type named ‘type’ in ‘struct std::thread::_Invoker<std::tuple<void (*)(float*), float (*)[10]> >::__result<std::tuple<void (*)(float*), float (*)[10]> >’
  247 |  operator()()
      |  ^~~~~~~~

我以为我忘记了dma线程的一些ctor参数，但事实并非如此。下面是我的代码：

#include <iostream>
#include <cstdint>
#include <thread>

void dma_service(bool*, float*, float*, float*);
void live_audio(float*);
void live_draw(float*);


    
    constexpr int d_size = 10;
    float data[d_size];
    float pwm_cp[d_size];
    float fft_cp[d_size];
    bool drdy = 0;
    std::thread dma_thrd(dma_service, &drdy, &data, &pwm_cp, &fft_cp);
    std::thread audio_thrd(live_draw, &pwm_cp);


int main(){
    for (auto i = 0; i < d_size; i++)
    {
        data[i] = 0;
        pwm_cp[i] = 0;
    }
    printf("Hello, World!\n");
    drdy = 1;

    int k = 3;
    while(k-->0)
    {
        audio_thrd.join();
        printf("\n\n\n");
        
        for (auto i = 0; i < d_size; i++)
        {
            data[i]++;
        }
        drdy = 1;
    }

}

void dma_service(bool* flag, float* d_array, float* pwm_array, float* fft_array) {
    while (!flag) ;

    for (auto i = 0; i < d_size; i++)
    {
        pwm_array[i] = d_array[i];
        fft_array[i] = d_array[i];
    }

    flag = 0;

}

void live_audio(float* pwm_array) {
    dma_thrd.join();
    printf("Audio Array:\n");
    for (auto i = 0; i < d_size; i++)
    {
        printf("PWM_CP[%d] %f\n", i, pwm_array[i]);
    }
}

我对lambda函数没有太多的经验，所以我不知道这里是否需要它们（并且很乐意避免它们）。谢谢你的建议！

Answer 1

给定浮点数据[d_size]；，&data被推导为“指向数组的指针”类型。与float*参数不匹配。

您可以简单地省略&，那么数组到指针的衰减就会开始，类型就会匹配。

std::thread dma_thrd(dma_service, &drdy, data, pwm_cp, fft_cp);
std::thread audio_thrd(live_draw, pwm_cp);

或者传递第一个元素的地址：

std::thread dma_thrd(dma_service, &drdy, &data[0], &pwm_cp[0], &fft_cp[0]);
std::thread audio_thrd(live_draw, &pwm_cp[0]);

注意，这不足以使程序工作。我发现它缺少线程同步，在全局范围内启动线程是有问题的，因为它们将在main()之前启动。此外，循环中的join()也不能正常工作，因为join()只能执行一次。