OpenCL将输入映射到输出

Question

我有一个大小为14的输入[1,2,3,4,1,3,3,5,1,2,5,4,1,3]，并将其用作内核输入。此输入将从给定的输出缓冲区索引开始复制到输出缓冲区。这项工作将在单个工作项位置进行（出于实验目的）。然后，我使用格式(input_int --> work_item_id --> output_int)和复制的output打印出映射结果。映射显示所有输入字符串（大小14）都已映射到output，但是复制的output在结果图像中仅将7个插槽打印为Resulting output。工作项的数量为7，但我打算将整个输入复制到工作项0的输出。这是怎么回事？

程序：

#!/usr/bin/env python3
#pragma OPENCL EXTENSION cl_khr_byte_addressable_store : enable
 #pragma OPENCL EXTENSION cl_khr_byte_addressable_store : enable
import pyopencl as cl
import numpy as np
import seq

# Write down our kernel as a multiline string.
kernel = """
    __kernel void dragon(
    const int N,
    __global char *AplusB,   

    __global char *output
    )
{   
    int idx = get_global_id(0); 


        if(idx == 0){ //A+B
           printf ("\\n mappings from input to output in kernel at idx %d",idx);  
            for (int i = 0; i <14; i++){
                     output[i]= AplusB[i];
                     int a = output[i];
                     int b =AplusB[i]; 

                     printf("\\n %d --> %d --> %d \\n",b,idx, a);
            }
        }
   }
"""
#declare constants
number_of_expansions = 4
total_probelem_size =7
resulting_str_size=62

# Step 1: Create a context.
# This will ask the user to select the device to be used.
context = cl.create_some_context()
# Create a queue to the device.
queue = cl.CommandQueue(context)
# Create the program.
program = cl.Program(context, kernel).build()
# Create the input string
AplusB  = np.array([1,2,3,4,1,3,3,5,1,2,5,4,1,3], dtype = np.int8)


#prepare out buffers
output = np.empty(total_probelem_size).astype(np.int8)
output.fill(0)

 # Create the memory on the device to put the result into.
expanded_output = cl.Buffer(context, cl.mem_flags.WRITE_ONLY, output.nbytes)


# Send the data to the guest memory.

mf = cl.mem_flags
AplusBBuf = cl.Buffer(context, mf.READ_ONLY | mf.COPY_HOST_PTR, hostbuf=AplusB)



# Initiate the kernel.
dragon = program.dragon
dragon.set_scalar_arg_dtypes([np.int32, None,None])

global_work_size = total_probelem_size
# Execute C = A * B.
dragon(queue, (global_work_size,), None,total_probelem_size,AplusBBuf,expanded_output)
# Wait for the queue to be completely processed.
queue.finish()
# Read the array from the device.

cl.enqueue_copy(queue, output, expanded_output).wait()
print("----------- Resulting output -------------------")
print (output)

结果：

Answer 1

expanded_output和output的大小为7个元素（total_probelem_size）。因此，从output[i=7]到output[i=13]的内核将写入无效的内存位置。

由于output数组仅包含7个元素，因此仅打印7个元素。