深度学习operation: matmul之NV shader实现

最近工作需要写shader，决定开个新坑，我写过的shader都会在后续文章公开。
写shader还是有必要了解一下NV GPU的硬件建构，至少知道device代码运行时的物理地址映射，CPU程序运行时有堆和栈的概念，而NV GPU 运行时是不同的，例如CPU c++ 函数内局部变量放在栈里，但NV GPU c++函数局部变量放在相应的SP寄存器里，特别是要写出高性能shader，往往需要对针对特定硬件架构做优化。
此实现利用了share memory做加速。
下面是工程结构：

build.sh

nvcc -x cu matmul.cu -o matmul

kernels.h

#ifndef KERNELS_H
#define KERNELS_H
#include "kernels.cu"
#endif

kernels.cu

#include <cuda_runtime.h>
#include <device_launch_parameters.h>

#define BLOCK_SIZE 16

// Matrices are stored in row-major order:
// M(row, col) = *(M.elements + row * M.stride + col)
typedef struct {
    int width;
    int height;
    int stride; 
    float* elements;
} Matrix;

// Get a matrix element
__device__ float GetElement(const Matrix A, int row, int col)
{
    return A.elements[row * A.stride + col];
}

// Set a matrix element
__device__ void SetElement(Matrix A, int row, int col,
                           float value)
{
    A.elements[row * A.stride + col] = value;
}

// Get the BLOCK_SIZExBLOCK_SIZE sub-matrix Asub of A that is
// located col sub-matrices to the right and row sub-matrices down
// from the upper-left corner of A
 __device__ Matrix GetSubMatrix(Matrix A, int row, int col) 
{
	//row = 1, col = 2;
    Matrix Asub;
    Asub.width    = BLOCK_SIZE;
    Asub.height   = BLOCK_SIZE;
    Asub.stride   = A.stride;
    Asub.elements = &A.elements[A.stride * BLOCK_SIZE * row
                                         + BLOCK_SIZE * col];
    return Asub;
}

// Thread block size


// Forward declaration of the matrix multiplication kernel
__global__ void MatMulKernel(const Matrix, const Matrix, Matrix);


// Matrix multiplication kernel called by MatMul()
 __global__ void MatMulKernel(const Matrix A, const Matrix B, Matrix C)
{
    // Block row and column
    int blockRow = blockIdx.y;
    int blockCol = blockIdx.x;

    // Each thread block computes one sub-matrix Csub of C
    Matrix Csub = GetSubMatrix(C, blockRow, blockCol);

    // Each thread computes one element of Csub
    // by accumulating results into Cvalue
    float Cvalue = 0;

    // Thread row and column within Csub
    int row = threadIdx.y;
    int col = threadIdx.x;

    // Loop over all the sub-matrices of A and B that are
    // required to compute Csub
    // Multiply each pair of sub-matrices together
    // and accumulate the results
    for (int m = 0; m < (A.width / BLOCK_SIZE); ++m) {

        // Get sub-matrix Asub of A
        Matrix Asub = GetSubMatrix(A, blockRow, m);

        // Get sub-matrix Bsub of B
        Matrix Bsub = GetSubMatrix(B, m, blockCol);

        // Shared memory used to store Asub and Bsub respectively
        __shared__ float As[BLOCK_SIZE][BLOCK_SIZE];
        __shared__ float Bs[BLOCK_SIZE][BLOCK_SIZE];

        // Load Asub and Bsub from device memory to shared memory
        // Each thread loads one element of each sub-matrix
        As[row][col] = GetElement(Asub, row, col);
        Bs[row][col] = GetElement(Bsub, row, col);

        // Synchronize to make sure the sub-matrices are loaded
        // before starting the computation
        __syncthreads();
        // Multiply Asub and Bsub together
        for (int e = 0; e < BLOCK_SIZE; ++e)
            Cvalue += As[row][e] * Bs[e][col];

        // Synchronize to make sure that the preceding
        // computation is done before loading two new
        // sub-matrices of A and B in the next iteration
        __syncthreads();
    }

    // Write Csub to device memory
    // Each thread writes one element
    SetElement(Csub, row, col, Cvalue);
}

matmul.cu

#include <cstdio>
#include <cstdlib>
#include <iostream>
#include <ctime>
#include <cuda_runtime.h>
#include <device_launch_parameters.h>

#include "kernels.h"

#define BLOCK_SIZE 16

int GenerateRandInt(int min, int max, int counter)
{
    srand(((int)time(0)) + counter);
    return (rand()%(max - min + 1) + min);
}


// Matrix multiplication - Host code
// Matrix dimensions are assumed to be multiples of BLOCK_SIZE
void MatMul(const Matrix A, const Matrix B, Matrix C)
{
    // Load A and B to device memory
    Matrix d_A;
    d_A.width = d_A.stride = A.width; d_A.height = A.height;
    size_t size = A.width * A.height * sizeof(float);
    cudaMalloc(&d_A.elements, size);
    cudaMemcpy(d_A.elements, A.elements, size,
               cudaMemcpyHostToDevice);
    Matrix d_B;
    d_B.width = d_B.stride = B.width; d_B.height = B.height;
    size = B.width * B.height * sizeof(float);
    cudaMalloc(&d_B.elements, size);
    cudaMemcpy(d_B.elements, B.elements, size,
    cudaMemcpyHostToDevice);

    // Allocate C in device memory
    Matrix d_C;
    d_C.width = d_C.stride = C.width; d_C.height = C.height;
    size = C.width * C.height * sizeof(float);
    cudaMalloc(&d_C.elements, size);

    // Invoke kernel
    dim3 dimBlock(BLOCK_SIZE, BLOCK_SIZE);
    dim3 dimGrid(B.width / dimBlock.x, A.height / dimBlock.y);
    MatMulKernel<<<dimGrid, dimBlock>>>(d_A, d_B, d_C);

    // Read C from device memory
    cudaMemcpy(C.elements, d_C.elements, size,
               cudaMemcpyDeviceToHost);

    // Free device memory
    cudaFree(d_A.elements);
    cudaFree(d_B.elements);
    cudaFree(d_C.elements);
}

int main(int argc, char** argv)
{
    int size_A = BLOCK_SIZE*2*BLOCK_SIZE*3;
    float matrix_A[size_A];
    std::cout<<"AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA"<<std::endl;
    for(int i=0; i<size_A; i++)
    {   
        matrix_A[i] = (float)GenerateRandInt(1,100,i);
        std::cout<<matrix_A[i]<<", ";
    }
    std::cout<<std::endl;
    Matrix A;
    A.height = BLOCK_SIZE*2;
    A.width = BLOCK_SIZE*3;
    A.elements = matrix_A;

    int size_B = BLOCK_SIZE*3*BLOCK_SIZE*4;
    float matrix_B[size_B];
    std::cout<<"BBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBB"<<std::endl;
    for(int i=0; i<size_B; i++)
    {   
        matrix_B[i] = (float)GenerateRandInt(1,100,i);
        std::cout<<matrix_B[i]<<", ";
    }
    std::cout<<std::endl;
    Matrix B;
    B.height = BLOCK_SIZE*3;
    B.width = BLOCK_SIZE*4;
    B.elements = matrix_B;

    int size_C = BLOCK_SIZE*2*BLOCK_SIZE*4;
    float matrix_C[size_C];
    Matrix C;
    C.height = BLOCK_SIZE*3;
    C.width = BLOCK_SIZE*4;
    C.elements = matrix_C;

    MatMul(A,B,C);

    std::cout<<"CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC"<<std::endl;
    for(int i=0; i<size_C; i++)
    {   
        std::cout<<C.elements[i]<<", ";
    }
    std::cout<<std::endl;

    return 0;
}

本文地址：https://blog.csdn.net/qq_33345917/article/details/107578989