Hi genaganna,
thank you for helping me.
But I can't post the kernel i'm still working. I'm not allowed to make it public. So I have a very simple kernel. its nonsence, but it has the same problem. i have a long loop, so its possible to see the cpu load.
Kernel:
__kernel void test1(__global float * a,__global float * b,__global float * c,__global const int *d){
int gid=get_global_id(0);
int gsi=get_global_size(0);
int step=d[0]/get_global_size(0);
for(int i=0;i<step;i++){
for(int j=0;j<1000000;j++){
c[gid+i*gsi]+=a[gid+i*gsi]+b[gid+i*gsi];
}
}
}
Host:
#include <cstdio>
#include <cstdlib>
#include <fstream>
#include <iostream>
#include <vector>
#include <iterator>
#define __CL_ENABLE_EXCEPTIONS
#include <CL\cl.hpp>
#include <ctime>
int main(int argc, char** argv){
cl_int error;
std::string buildlog;
cl::Context context;
cl::Program program;
std::vector<cl::Device> devices;
try{
//get CL platform info
std::vector<cl::Platform> platforms;
cl::Platform::get(&platforms);
cl_context_properties platform=NULL;
std::vector<cl::Platform>::iterator i;
if(platforms.size() > 0){
for(i = platforms.begin(); i != platforms.end(); ++i){
platform=((cl_context_properties)(*i)());
if(!strcmp((*i).getInfo<CL_PLATFORM_VENDOR>().c_str(), "Advanced Micro Devices, Inc."))break;
}
}
cl_context_properties cps[3] = { CL_CONTEXT_PLATFORM, platform, 0 };
cl_context_properties *cprops =(platform==NULL) ? NULL : cps;
//Creating CL Device;
context=cl::Context(CL_DEVICE_TYPE_CPU,cprops,NULL,NULL,&error);
//getting Device List
devices=context.getInfo<CL_CONTEXT_DEVICES>();
//creating Commandqueue
cl::CommandQueue queue=cl::CommandQueue(context,devices[0]);
//Reading CL Programm from file
std::ifstream file("BastelKernel.cl"); //Kernelname
std::string prog(std::istreambuf_iterator<char>(file),(std::istreambuf_iterator<char>()));
cl::Program::Sources source(1,std::make_pair(prog.c_str(), prog.length()));
//Building CL Programm for Device
program=cl::Program(context,source,&error);
program.build(devices);
//finally Kernels:
cl::Kernel kernel1=cl::Kernel(program,"test1",&error);
//Hostmemory
cl_int wsize=8192;
cl_int worksize;
int cpun;
devices[0].getInfo(CL_DEVICE_MAX_WORK_GROUP_SIZE,&worksize);
devices[0].getInfo(CL_DEVICE_MAX_COMPUTE_UNITS,&cpun);
std::cout<<"Max Compute Units: "<<cpun<<std::endl;
cl_float *a=new cl_float[wsize];
cl_float *b=new cl_float[wsize];
cl_float *c=new cl_float[wsize];
cl_int *d=new cl_int[1]; d[0]=wsize;
//initialing OpenCL Buffer(MemoryObjects)
cl::Buffer CL1=cl::Buffer(context,CL_MEM_READ_ONLY |CL_MEM_USE_HOST_PTR,sizeof(a[0]) * wsize,a,&error);
cl::Buffer CL2=cl::Buffer(context,CL_MEM_READ_ONLY |CL_MEM_USE_HOST_PTR,sizeof(b[0]) * wsize,b,&error);
cl::Buffer CL3=cl::Buffer(context,CL_MEM_READ_WRITE|CL_MEM_USE_HOST_PTR,sizeof(c[0]) * wsize,c,&error);
cl::Buffer CL4=cl::Buffer(context,CL_MEM_READ_ONLY |CL_MEM_USE_HOST_PTR,sizeof(c[0]) * wsize,d,&error);
//set Hostmemory
for(int i=0;i<wsize;i++){
a=i+0.1;
b=wsize-i-0.1;
c=0;
}
//set Kernel Arguments
kernel1.setArg(0,CL1);
kernel1.setArg(1,CL2);
kernel1.setArg(2,CL3);
kernel1.setArg(3,CL4);
//Running Kernel
clock_t time;
time=clock();
queue.finish();
queue.enqueueNDRangeKernel(kernel1,cl::NullRange,cl::NDRange(worksize,1,1),cl::NDRange(worksize,1,1),NULL,NULL); //queue.enqueueNDRangeKernel(kernelname,cl::NullRange,cl::NDRange(arraylänge),cl::NDRange(1,1),NULL,NULL);
queue.enqueueReadBuffer (CL3,CL_TRUE,0,sizeof(c[0])*wsize,c);
queue.finish();
time=clock()-time;
//Ausgabe
std::cout<<std::endl<<"Ergebnis OCL: "<<std::endl<<"";
for(int i=0;i<wsize;i+=100){
std::cout<<c<<" ";
}
std::cout<<std::endl;
std::cout<<std::endl<<"Zeit OCL: "<<time<<"ms"<<std::endl<<std::endl;
delete[] a,b,c;
}catch(cl::Error& error){
std::cout<<"OpenCL-Error: "<<error.what()<<"("<<error.err()<<")"<<std::endl<<std::endl;
}
std::cout<<std::endl<<"________________________________________________________________________________"<<std::endl;
std::cout<<"Buildlog:"<<std::endl;
buildlog=program.getBuildInfo<CL_PROGRAM_BUILD_LOG>(devices[0]);
std::cout<<buildlog<<std::endl;
std::cout<<"________________________________________________________________________________"<<std::endl;;
return 0;
}
it just runs on one core, 100% load, others are 0% load.
On GPU ist runs with gpu-load 99%...
MFG
Sheeep
