cuda内核是否受每个线程/块的内存使用限制

Question

我有一个内核代码whixh正确执行 可运行代码

__global__ static void CalcSTLDistance_Kernel(Integer ComputeParticleNumber)
{
    //const Integer TID = CudaGetTargetID();
    const Integer ID  =CudaGetTargetID(); 
    /*if(ID >= ComputeParticleNumber)
    {
        return ;
    }*/
    CDistance NearestDistance;
    Integer NearestID = -1;
    NearestDistance.Magnitude = 1e8;
    NearestDistance.Direction.x = 0;
    NearestDistance.Direction.y = 0;
    NearestDistance.Direction.z = 0;//make_Scalar3(0,0,0);
    //if(c_daOutputParticleID[ID] < -1)
    //{
    //  c_daSTLDistance[ID] = NearestDistance;
    //  c_daSTLID[ID] = NearestID;
    //  return;
    //}

    //Scalar3 TargetPosition = c_daParticlePosition[ID];

    Integer TriangleID;     
    Integer CIDX, CIDY, CIDZ;
    Integer CID = GetCellID(&CONSTANT_BOUNDINGBOX,&c_daParticlePosition[ID],CIDX, CIDY, CIDZ);
    if(CID >=0 && CID < c_CellNum)
    {
        //Integer Range = 1;
        for(Integer k = -1; k <= 1; ++k)
        {
            for(Integer j = -1; j <= 1; ++j)
            {
                for(Integer i = -1; i <= 1; ++i)
                {
                    Integer MCID = GetCellID(&CONSTANT_BOUNDINGBOX,CIDX +i, CIDY + j,CIDZ + k);
                    if(MCID < 0 || MCID >= c_CellNum)
                    {
                        continue;
                    }
                    unsigned int TriangleNum = c_daCell[MCID].m_TriangleNum;
                    for(unsigned int l = 0; l < TriangleNum; ++l)
                    {
                        TriangleID = c_daCell[MCID].m_TriangleID[l];
                        /*if(c_daTrianglesParameters[c_daTriangles[TriangleID].ModelIDNumber].isDrag)
                        {
                            continue;
                        }*/

                        if( TriangleID >= 0 && TriangleID < c_TriangleNum && TriangleID != NearestID)// No need to calculate again for the same triangle
                        {
                        CDistance Distance ;
                            Distance.Magnitude = CalcDistance(&c_daTriangles[TriangleID], &c_daParticlePosition[ID], &Distance.Direction);
                            if(Distance.Magnitude < NearestDistance.Magnitude)
                            {
                                NearestDistance = Distance;
                                NearestID = TriangleID;
                            }
                        }
                    }   
                }
            }
        }
    }
    c_daSTLDistance[ID] = NearestDistance;
    c_daSTLID[ID] = NearestID;
}

当我添加任何基本变量或执行任何检查操作时，它会产生未知错误，并且在检查cuda-memcheck时，它会建议内存读取错误。

这里在更改的代码中，我试图检查先前计算的部分并尝试跳过冗余计算。为此，我尝试在数组中执行基本检查操作，但它抛出内存错误

错误提升代码

__global__ static void CalcSTLDistance_Kernel(Integer ComputeParticleNumber)
{
    //const Integer TID = CudaGetTargetID();
    const Integer ID  =CudaGetTargetID(); 
    /*if(ID >= ComputeParticleNumber)
    {
        return ;
    }*/
    CDistance NearestDistance;
    Integer NearestID = -1;
    NearestDistance.Magnitude = 1e8;
    NearestDistance.Direction.x = 0;
    NearestDistance.Direction.y = 0;
    NearestDistance.Direction.z = 0;//make_Scalar3(0,0,0);
    //if(c_daOutputParticleID[ID] < -1)
    //{
    //  c_daSTLDistance[ID] = NearestDistance;
    //  c_daSTLID[ID] = NearestID;
    //  return;
    //}

    //Scalar3 TargetPosition = c_daParticlePosition[ID];

    Integer TriangleID;     
    Integer CIDX, CIDY, CIDZ;
    Integer CID = GetCellID(&CONSTANT_BOUNDINGBOX,&c_daParticlePosition[ID],CIDX, CIDY, CIDZ);
    int len=0;
    int td[100];
    for(int m=0;m<100;m++)
    {
      td[m]=-1;

    }
    if(CID >=0 && CID < c_CellNum)
    {
        //Integer Range = 1;
        for(Integer k = -1; k <= 1; ++k)
        {
            for(Integer j = -1; j <= 1; ++j)
            {
                for(Integer i = -1; i <= 1; ++i)
                {
                    Integer MCID = GetCellID(&CONSTANT_BOUNDINGBOX,CIDX +i, CIDY + j,CIDZ + k);
                    if(MCID < 0 || MCID >= c_CellNum)
                    {
                        continue;
                    }
                    unsigned int TriangleNum = c_daCell[MCID].m_TriangleNum;



                     bool flag = false;

                       //len=len+TriangleNum ;
                     for(unsigned int l = 0; l < TriangleNum; ++l)
                     {
                        TriangleID = c_daCell[MCID].m_TriangleID[l]; 
                        //tem[l] = c_daCell[MCID].m_TriangleID[l];

                        for(int m=0;m<100;m++)
                        {
                            if(TriangleID ==td[m])
                            {
                                flag= true;



                            }
                            if(flag == true)
                            break;


                        }



                        if(flag == true)
                            continue;
                        else 
                        {
                            td[len] = TriangleID;
                            len= len+1;

                            if( TriangleID >= 0 && TriangleID < c_TriangleNum && TriangleID != NearestID)// No need to calculate again for the same triangle
                             {
                              CDistance Distance ;
                              Distance.Magnitude = CalcDistance(&c_daTriangles[TriangleID], &c_daParticlePosition[ID], &Distance.Direction);
                              if(Distance.Magnitude < NearestDistance.Magnitude)
                                {
                                NearestDistance = Distance;
                                NearestID = TriangleID;
                                }
                             }

                        }
                     }

                }
            }
        }
    }
    c_daSTLDistance[ID] = NearestDistance;
    c_daSTLID[ID] = NearestID;
}

每当我尝试添加任何代码时都会出现这个问题，因此我怀疑这块内核不允许我添加任何进一步的代码，因为内存过度使用。 每个块或线程是否有任何内存违规规则??

如何查找每个内核的总内存使用情况？有什么办法吗？