9 Replies Latest reply on Feb 26, 2013 10:44 PM by himanshu.gautam

    Need help with this code

    notooth

      Hello,

       

      I have a code and the output like this:

      __kernel void Generate21(__global ulong* out)

      {

                int num = get_global_id(0);

                out[num] = (ulong)0x0123456789abcdef;

      }

       

      cl_mem cl_Matrix = clCreateBuffer(context, CL_MEM_READ_WRITE, 2*64*8, NULL, NULL);

      cl_kernel Generate21 = clCreateKernel(program,"Generate21", NULL);

      status = clSetKernelArg(Generate21, 0, sizeof(cl_mem), (void *)&cl_Matrix);

      *global_work_size = 2;

      status = clEnqueueNDRangeKernel(commandQueue, Generate21, 1, NULL, global_work_size, NULL, 0, NULL, NULL);

      __int64 Matrix[2][64];

      status = clEnqueueReadBuffer(commandQueue, cl_Matrix, CL_TRUE, 0, 2*64*8, Matrix, 0, NULL, NULL);

      for (char k=1; k>=0; k--)

                {

                          for (char i=63; i>=0; i--)

                          {

                                    for (char j=63; j>=0; j--)

                                              printf("%d", (Matrix[k][i]>>j)&1);

                                    printf("\n");

                          }

                          printf("\n");

                }

       

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

       

       

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000100100011010001010110011110001001101010111100110111101111

      0000000100100011010001010110011110001001101010111100110111101111

       

       

      Can you help me fix the code to get the output like this, in 1 command to execute the kernel:

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000100100011010001010110011110001001101010111100110111101111

       

       

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000000000000000000000000000000000000000000000000000000000000

      0000000100100011010001010110011110001001101010111100110111101111

       

        • Re: Need help with this code
          himanshu.gautam

          What is __int64? The right thing to use is "cl_ulong".

           

          Can you confirm if the issue reported http://devgurus.amd.com/message/1287346#1287346 is solved?

          Also, I am not sure if "%d" would print a 64-bit integer. Can you check?

          In case of doubt, use std::cout to print.

          Also, Please tell which platform you are working on. Is it 32-bit (or) 64-bit?

           

          Also, using a "char" for a for loop index looks weird. Can you use an "int" instead?

           

          Also, Your print messages are not separated by even a space.

          Nobody can make sense out of this kind of messages.

           

          Please tell us what is the expectation mismatch here. What is going wrong?

          Why cant you just print the output of the kernel and see if the value is set to whatever you are setting the kernel.

            • Re: Need help with this code
              notooth

              __int64 is an 8 bytes data type. I think "cl_ulong" is a data type created on GPU, and "__int64" is a data type created on host. I replied your answer, please have a look http://devgurus.amd.com/message/1287346#1287346

               

              "%d" would not print a 64-bit integer. In my code, "%d" printed bits of a 64-bit integer. I used "char" for a loop index to safe resource of the host. But "int" is also ok for me.

                • Re: Need help with this code
                  himanshu.gautam

                  notooth.

                  Your understanding is wrong.

                  "ulong" inside OpenCL kernel represents 64-bit unsigned integer - Please refer Chapter 6 of OpenCL spec.

                  "cl_ulong" is the host side data-type that corresponds to "ulong" in the kernel.

                   

                  i.e. If you wanted to create an array of 64-bit integers, you need to use "cl_ulong" in your host code

                  and access it as a "ulong *" pointer in the kernel code.

                  Hope this is clear.

              • Re: Need help with this code
                german

                notooth wrote:

                Can you help me fix the code to get the output like this, in 1 command to execute the kernel:

                The simplest way should be

                __kernel  void Generate21(__global ulong* out)

                {

                  int num = get_global_id(0);

                    out[num * 64] = (ulong)0x0123456789abcdef;

                }

                  • Re: Need help with this code
                    himanshu.gautam

                     

                    __kernel  void Generate21(__global ulong* out)

                    {

                      int num = get_global_id(0);

                        out[num * 64] = (ulong)0x0123456789abcdef;

                    }

                     

                    No. He actually needs to spin a FOR loop for 64 times. (his global work size is actually 2)

                    The code snippet above will set two items that are 64 ulongs apart.

                      • Re: Need help with this code
                        german

                        himanshu.gautam wrote:

                         

                         

                        __kernel  void Generate21(__global ulong* out)

                        {

                          int num = get_global_id(0);

                            out[num * 64] = (ulong)0x0123456789abcdef;

                        }

                         

                        No. He actually needs to spin a FOR loop for 64 times. (his global work size is actually 2)

                        The code snippet above will set two items that are 64 ulongs apart.

                        notooth didn't ask how to clear the rest of memory. The question was how to make the output like that. So that is the answer. 2 values - 2 locations. Nothing less or nothing more.

                    • Re: Need help with this code
                      himanshu.gautam

                      Notooth,

                      You are spawning only 2 workitems (global size)!

                      And, I dont see any for loops inside your kernel.

                      So, effectively you are setting only 2 elements in the array.

                      Also, I am not sure if global work size of 2 will be accepted... Just check the status of the kernel call as well.

                      • Re: Need help with this code
                        notooth

                        Thanks for all the replies.