为什么我的CUDA内核会因不同的数据集大小而崩溃（未指定的启动失败）？

我有一个内核来计算矩阵的不同元素，基于它们的位置（对角线或非对角线）。计算大小矩阵时，内核按预期工作：

• 14 x 14（我知道这很小，并且没有正确使用GPU资源，但这纯粹是为了测试目的，以确保结果正确）
• 118 x 118，以及
• 300 x 300

然而，当我试图计算一个大小为2383x2383的矩阵时，内核崩溃了。具体来说，cudaMemcpy（）行上会抛出错误“Unspecified launch failure”，以将结果从设备返回到主机。通过研究，我了解到这种错误通常发生在越界内存访问的情况下（例如，在数组中），但是，我不知道它适用于前面三种情况，但不适用于2383 x 2383情况。内核代码如下所示：

__global__ void createYBus(float *R, float *X, float *B, int numberOfBuses, int numberOfBranches, int *fromBus, int *toBus, cuComplex *y)
{
    int rowIdx = blockIdx.y*blockDim.y + threadIdx.y;
    int colIdx = blockIdx.x*blockDim.x + threadIdx.x;
    int index = rowIdx*numberOfBuses + colIdx;
    if (rowIdx<numberOfBuses && colIdx<numberOfBuses)
    {
        for (int i=0; i<numberOfBranches; ++i)
        {
            if (rowIdx==fromBus[i] && colIdx==fromBus[i]) { //diagonal element
                y[index] = cuCaddf(y[index], make_cuComplex((R[i]/((R[i]*R[i])+(X[i]*X[i]))), (-(X[i]/((R[i]*R[i])+(X[i]*X[i])))+ (B[i]/2))));
            }
            if (rowIdx==toBus[i] && colIdx==toBus[i]) { //diagonal element
                y[index] = cuCaddf(y[index], make_cuComplex((R[i]/((R[i]*R[i])+(X[i]*X[i]))), (-(X[i]/((R[i]*R[i])+(X[i]*X[i])))+ (B[i]/2))));
            }
            if (rowIdx==fromBus[i] && colIdx==toBus[i]) { //off-diagonal element
                y[index] = make_cuComplex(-(R[i]/((R[i]*R[i])+(X[i]*X[i]))), X[i]/((R[i]*R[i])+(X[i]*X[i])));
            }
            if (rowIdx==toBus[i] && colIdx==fromBus[i]) { //off-diagonal element
                y[index] = make_cuComplex(-(R[i]/((R[i]*R[i])+(X[i]*X[i]))), X[i]/((R[i]*R[i])+(X[i]*X[i])));
            }
        }
    }
}

其中cudaStatN为cudaError\t类型，用于捕获错误。最后四次分配是在代码后面完成的，并且是针对另一个内核的。然而，这些分配是在调用相关内核之前完成的

发射参数如下：

cudaStat1 = cudaMalloc((void**)&dev_fromBus, numLines*sizeof(int));
cudaStat2 = cudaMalloc((void**)&dev_toBus, numLines*sizeof(int));
cudaStat3 = cudaMalloc((void**)&dev_R, numLines*sizeof(float));
cudaStat4 = cudaMalloc((void**)&dev_X, numLines*sizeof(float));
cudaStat5 = cudaMalloc((void**)&dev_B, numLines*sizeof(float));
cudaStat6 = cudaMalloc((void**)&dev_y, numberOfBuses*numberOfBuses*sizeof(cuComplex));
cudaStat7 = cudaMalloc((void**)&dev_Pd, numberOfBuses*sizeof(float));
cudaStat8 = cudaMalloc((void**)&dev_Qd, numberOfBuses*sizeof(float));
cudaStat9 = cudaMalloc((void**)&dev_Vmag, numberOfBuses*sizeof(float));
cudaStat10 = cudaMalloc((void**)&dev_theta, numberOfBuses*sizeof(float));
cudaStat11 = cudaMalloc((void**)&dev_Peq, numberOfBuses*sizeof(float));
cudaStat12 = cudaMalloc((void**)&dev_Qeq, numberOfBuses*sizeof(float));
cudaStat13 = cudaMalloc((void**)&dev_Peq1, numberOfBuses*sizeof(float));
cudaStat14 = cudaMalloc((void**)&dev_Qeq1, numberOfBuses*sizeof(float));
...
...
cudaStat15 = cudaMalloc((void**)&dev_powerMismatch, jacSize*sizeof(float));
cudaStat16 = cudaMalloc((void**)&dev_jacobian, jacSize*jacSize*sizeof(float));
cudaStat17 = cudaMalloc((void**)&dev_stateVector, jacSize*sizeof(float));
cudaStat18 = cudaMalloc((void**)&dev_PQindex, jacSize*sizeof(int));

dim3 dimBlock(16, 16); //number of threads 
dim3 dimGrid((numberOfBuses+15)/16, (numberOfBuses+15)/16);  //number of blocks

//launch kernel once data has been copied to GPU
createYBus<<<dimGrid, dimBlock>>>(dev_R, dev_X, dev_B, numberOfBuses, numLines, dev_fromBus, dev_toBus, dev_y);

//copy results back to CPU
cudaStat6 = cudaMemcpy(y_bus, dev_y, numberOfBuses*numberOfBuses*sizeof(cuComplex), cudaMemcpyDeviceToHost);
if (cudaStat6 != cudaSuccess) {
    cout<<"Device memcpy failed"<<endl;
    cout<<cudaGetErrorString(cudaStat6)<<endl;
    return 1;
}

---

我知道我的内核不是最高效的，因为有很多全局内存访问。为什么这个较大的矩阵的内核会崩溃？是否有我丢失的越界数组访问？非常感谢您的帮助。

我试图用下面的完整示例重现您的代码。代码编译、运行时没有错误

#include "cuda_runtime.h"
#include "device_launch_parameters.h"

#include <stdio.h>

#include "cuComplex.h"

__global__ void createYBus(float *R, float *X, float *B, int numberOfBuses, int numberOfBranches, int *fromBus, int *toBus, cuComplex *y)
{
    int rowIdx = blockIdx.y*blockDim.y + threadIdx.y;
    int colIdx = blockIdx.x*blockDim.x + threadIdx.x;
    int index = rowIdx*numberOfBuses + colIdx;
    if (rowIdx<numberOfBuses && colIdx<numberOfBuses)
    {
        for (int i=0; i<numberOfBranches; ++i)
        {
            if (rowIdx==fromBus[i] && colIdx==fromBus[i]) { //diagonal element
                y[index] = cuCaddf(y[index], make_cuComplex((R[i]/((R[i]*R[i])+(X[i]*X[i]))), (-(X[i]/((R[i]*R[i])+(X[i]*X[i])))+ (B[i]/2))));
            }
            if (rowIdx==toBus[i] && colIdx==toBus[i]) { //diagonal element
                y[index] = cuCaddf(y[index], make_cuComplex((R[i]/((R[i]*R[i])+(X[i]*X[i]))), (-(X[i]/((R[i]*R[i])+(X[i]*X[i])))+ (B[i]/2))));
            }
            if (rowIdx==fromBus[i] && colIdx==toBus[i]) { //off-diagonal element
                y[index] = make_cuComplex(-(R[i]/((R[i]*R[i])+(X[i]*X[i]))), X[i]/((R[i]*R[i])+(X[i]*X[i])));
            }
            if (rowIdx==toBus[i] && colIdx==fromBus[i]) { //off-diagonal element
                y[index] = make_cuComplex(-(R[i]/((R[i]*R[i])+(X[i]*X[i]))), X[i]/((R[i]*R[i])+(X[i]*X[i])));
            }
        }
    }
}


int main ()
{
    int numLines = 32 ;
    int numberOfBuses = 2383 ;

    int* dev_fromBus, *dev_toBus;
    float *dev_R, *dev_X, *dev_B;
    cuComplex* dev_y ; 

    cudaMalloc((void**)&dev_fromBus, numLines*sizeof(int));
    cudaMalloc((void**)&dev_toBus, numLines*sizeof(int));
    cudaMalloc((void**)&dev_R, numLines*sizeof(float));
    cudaMalloc((void**)&dev_X, numLines*sizeof(float));
    cudaMalloc((void**)&dev_B, numLines*sizeof(float));
    cudaMalloc((void**)&dev_y, numberOfBuses*numberOfBuses*sizeof(cuComplex));

    dim3 dimBlock(16, 16); //number of threads 
    dim3 dimGrid((numberOfBuses+15)/16, (numberOfBuses+15)/16);  //number of blocks

    //launch kernel once data has been copied to GPU
    createYBus<<<dimGrid, dimBlock>>>(dev_R, dev_X, dev_B, numberOfBuses, numLines, dev_fromBus, dev_toBus, dev_y);

    cuComplex* y_bus = new cuComplex[numberOfBuses*numberOfBuses] ;

    //copy results back to CPU
    cudaError_t cudaStat6 = cudaMemcpy(y_bus, dev_y, numberOfBuses*numberOfBuses*sizeof(cuComplex), cudaMemcpyDeviceToHost);
    if (cudaStat6 != cudaSuccess) {
        printf ("failure : (%d) - %s\n", cudaStat6, ::cudaGetErrorString(cudaStat6)) ;
        return 1;
    }
    return 0 ;
}

#包括“cuda_runtime.h”
#包括“设备启动参数.h”
#包括
#包括“cuComplex.h”
__全局无效创建总线（float*R、float*X、float*B、int numberofbus、int numberofbranchs、int*fromBus、int*toBus、cuComplex*y）
{
int rowIdx=blockIdx.y*blockDim.y+threadIdx.y；
int colIdx=blockIdx.x*blockDim.x+threadIdx.x；
int index=rowIdx*numberofbus+colIdx；
如果（rowIdx解决了问题。结果显示WDDM TDR（超时检测恢复）已启用，延迟设置为2秒。这意味着如果内核执行时间超过2秒，驱动程序将崩溃并恢复。这适用于图形和渲染（用于GPU的一般用途）。但是，在这种情况下，TDR必须禁用me或延迟增加。通过将延迟增加到10秒，崩溃错误“未指定的启动失败”停止出现，内核执行仍像以前一样继续

TDR延迟（以及启用/禁用）可以通过Nsight监视器中的Nsight选项或通过注册表（HKEY\U LOCAL\U MACHINE\SYSTEM\CurrentControlSet\Control\GraphicsDrivers）完成-DWORDS Tdrdelay和Tdrlevel。
可能是内存分配大小与参数不匹配。是否也可以共享分配调用？是否确定“int index”的值，在你的内核中，是否正确？它不应该取决于块的尺寸吗？@FlorentDUGUET我编辑了这篇文章以包含分配调用。我知道全局内存很大，但有可能是我分配了太多内存吗？Cudamaloc（）没有任何错误调用。@Taro我正在启动一个包含2D块的2D网格。全局线程ID在x和y方向上计算，然后索引结合全局线程ID公式，以平展形式索引整个矩阵。我很好奇为什么要向下投票。我很乐意改进答案。这将与较小的情况一样有效。The numLines变量来自存储数据的文本文件中的行数。行数为2896。文本文件中的每列存储为一个数组（fromBus、toBus、R、X、B），然后复制到其分配的设备对应项（前缀为“dev”）。此外，cuComplex是标头中的CUDA数据类型“CuFix.h”附带工具包，所以没有必要执行。我将再次尝试NVISE，谢谢。（不确定下注）。不同的人认为是“为我工作”。"答案很差。几年前我第一次开始回答问题时就被告知了这一点。当然，这是一个意见问题，但这可能是否决票的一个原因。然而，人们普遍认为，要求调试帮助但未能提供MCVE的问题应该投票了结——这显然是错误的在SO帮助页面上显示。回答此类问题或试图这样做，也可能被一些人认为是不好的形式。相反，鼓励发布MCVE。如果您希望完成您在回答中所做的工作，一种可能的替代方法是在问题的注释中添加足够的描述性文本，建议应提供MCVEided，并在外部链接（如pastebin）中提供代码或完全工作的示例。如果您仔细阅读此处的问题，您会发现此类行为的示例。无论如何，我不想争论。如果我们不同意，也没关系。这是社区的本质。感谢@RobertCrovella的解释。问题是：为什么内核会因为t而崩溃他的更大的矩阵？是否有我缺少的越界数组访问？我试图在回答中表达的是，答案中提供的元素没有显示任何问题，而问题在其他地方。提供函数代码示例是为了说明我的说法。我不知道如何投票支持结束这个问题，但我会得到答案r如果被提议为danieljovan，则既没有提供MCVE，也没有将该答案标记为有效。再次感谢Robert。感谢分享，你是一个英雄！这正是我的问题，非常感谢。我想知道为什么通过Google找到这个答案如此困难？这通常适用于所有CUDA用例-特别是MATLAB。刚刚修复了我遇到的一个问题已经有好几天了！我已经为Tdrdelay创建了一个DWORD注册表项。我应该为Tdrlevel创建另一个注册表项吗？的值应该是多少
#include "cuda_runtime.h"
#include "device_launch_parameters.h"

#include <stdio.h>

#include "cuComplex.h"

__global__ void createYBus(float *R, float *X, float *B, int numberOfBuses, int numberOfBranches, int *fromBus, int *toBus, cuComplex *y)
{
    int rowIdx = blockIdx.y*blockDim.y + threadIdx.y;
    int colIdx = blockIdx.x*blockDim.x + threadIdx.x;
    int index = rowIdx*numberOfBuses + colIdx;
    if (rowIdx<numberOfBuses && colIdx<numberOfBuses)
    {
        for (int i=0; i<numberOfBranches; ++i)
        {
            if (rowIdx==fromBus[i] && colIdx==fromBus[i]) { //diagonal element
                y[index] = cuCaddf(y[index], make_cuComplex((R[i]/((R[i]*R[i])+(X[i]*X[i]))), (-(X[i]/((R[i]*R[i])+(X[i]*X[i])))+ (B[i]/2))));
            }
            if (rowIdx==toBus[i] && colIdx==toBus[i]) { //diagonal element
                y[index] = cuCaddf(y[index], make_cuComplex((R[i]/((R[i]*R[i])+(X[i]*X[i]))), (-(X[i]/((R[i]*R[i])+(X[i]*X[i])))+ (B[i]/2))));
            }
            if (rowIdx==fromBus[i] && colIdx==toBus[i]) { //off-diagonal element
                y[index] = make_cuComplex(-(R[i]/((R[i]*R[i])+(X[i]*X[i]))), X[i]/((R[i]*R[i])+(X[i]*X[i])));
            }
            if (rowIdx==toBus[i] && colIdx==fromBus[i]) { //off-diagonal element
                y[index] = make_cuComplex(-(R[i]/((R[i]*R[i])+(X[i]*X[i]))), X[i]/((R[i]*R[i])+(X[i]*X[i])));
            }
        }
    }
}


int main ()
{
    int numLines = 32 ;
    int numberOfBuses = 2383 ;

    int* dev_fromBus, *dev_toBus;
    float *dev_R, *dev_X, *dev_B;
    cuComplex* dev_y ; 

    cudaMalloc((void**)&dev_fromBus, numLines*sizeof(int));
    cudaMalloc((void**)&dev_toBus, numLines*sizeof(int));
    cudaMalloc((void**)&dev_R, numLines*sizeof(float));
    cudaMalloc((void**)&dev_X, numLines*sizeof(float));
    cudaMalloc((void**)&dev_B, numLines*sizeof(float));
    cudaMalloc((void**)&dev_y, numberOfBuses*numberOfBuses*sizeof(cuComplex));

    dim3 dimBlock(16, 16); //number of threads 
    dim3 dimGrid((numberOfBuses+15)/16, (numberOfBuses+15)/16);  //number of blocks

    //launch kernel once data has been copied to GPU
    createYBus<<<dimGrid, dimBlock>>>(dev_R, dev_X, dev_B, numberOfBuses, numLines, dev_fromBus, dev_toBus, dev_y);

    cuComplex* y_bus = new cuComplex[numberOfBuses*numberOfBuses] ;

    //copy results back to CPU
    cudaError_t cudaStat6 = cudaMemcpy(y_bus, dev_y, numberOfBuses*numberOfBuses*sizeof(cuComplex), cudaMemcpyDeviceToHost);
    if (cudaStat6 != cudaSuccess) {
        printf ("failure : (%d) - %s\n", cudaStat6, ::cudaGetErrorString(cudaStat6)) ;
        return 1;
    }
    return 0 ;
}