小编jxj*_*jxj的帖子

我有多个内核,它们以顺序方式启动,如下所示:

        clEnqueueNDRangeKernel(..., kernel1, ...);
        clEnqueueNDRangeKernel(..., kernel2, ...);
        clEnqueueNDRangeKernel(..., kernel3, ...);

并且,多个内核共享一个全局缓冲区.

现在,我通过在clEnqueueNDRangeKernel之后添加代码块来分析每个内核执行并总结它们以计算总执行时间:

        clFinish(cmdQueue);
        status = clGetEventProfilingInfo(...,&starttime,...);
        clGetEventProfilingInfo(...,&endtime,...);
        time_spent = endtime - starttime;

我的问题是如何通过一个clFinish一起分析三个内核？(比如在最后一次内核启动后添加一个clFinish()).

是的,我给每个clEnqueueNDRangeKernel不同的时间事件,并得到大的负数.详细信息:

clEnqueueNDRangeKernel(cmdQueue,...,&timing_event1);
clFinish(cmdQueue);
clGetEventProfilingInfo(timing_event1,CL_PROFILING_COMMAND_START,sizeof(cl_ulong),&starttime1,NULL);
clGetEventProfilingInfo(timing_event1,CL_PROFILING_COMMAND_END,sizeof(cl_ulong),&endtime1,NULL);
time_spent1 = endtime1 - starttime1;

clEnqueueNDRangeKernel(cmdQueue,...,&timing_event2);
clFinish(cmdQueue);
clGetEventProfilingInfo(timing_event2,CL_PROFILING_COMMAND_START,sizeof(cl_ulong),&starttime2,NULL);
clGetEventProfilingInfo(timing_event2,CL_PROFILING_COMMAND_END,sizeof(cl_ulong),&endtime2,NULL);
time_spent2 = endtime2 - starttime2;

clEnqueueNDRangeKernel(cmdQueue,...,&timing_event3);
clFinish(cmdQueue);
clGetEventProfilingInfo(timing_event3,CL_PROFILING_COMMAND_START,sizeof(cl_ulong),&starttime3,NULL);
clGetEventProfilingInfo(timing_event3,CL_PROFILING_COMMAND_END,sizeof(cl_ulong),&endtime3,NULL);
time_spent3 = endtime3 - starttime3;

time_spent_all_0 = time_spent1 + time_spent2 + time_spent3;
time_spent_all_1 = endtime3 - starttime1;

如果我有每个clFinish,所有分析值都是合理的,但time_spent_all_1大约是time_spent_all_0的2倍.如果我删除除最后一个clFinish之外的所有clFinish,则所有分析值都是不合理的.

感谢Eric Bainville,我得到了我想要的结果:通过一个clFinish分析多个clEnqueueNDRangeKernel.以下是我使用的最终代码:

clEnqueueNDRangeKernel(cmdQueue,...,&timing_event1);
clEnqueueNDRangeKernel(cmdQueue,...,&timing_event2);
clEnqueueNDRangeKernel(cmdQueue,...,&timing_event3);
clFinish(cmdQueue);

clGetEventProfilingInfo(timing_event1,CL_PROFILING_COMMAND_START,sizeof(cl_ulong),&starttime,NULL);
clGetEventProfilingInfo(timing_event3,CL_PROFILING_COMMAND_END,sizeof(cl_ulong),&endtime,NULL);
time_spent = endtime - starttime;

我正在为项目编写CMakeLists.txt,并遇到set_source_files_properties的问题.

原始工作表达式是:

set_source_files_properties (a.cpp PROPERTIES COMPILE_DEFINITIONS
    DIR1="/home/xxx/b.i")

然后我尝试添加更多COMPILE_DEFINITIONS,但是失败了.

尝试1:

set_source_files_properties (a.cpp PROPERTIES COMPILE_DEFINITIONS
    DIR1="/home/xxx/b.i" DIR2="/home/xxx/c.i" DIR3="/home/xxx/d.i")

尝试2:

set_source_files_properties (a.cpp PROPERTIES COMPILE_DEFINITIONS
    DIR1="/home/xxx/b.i")
set_source_files_properties (a.cpp PROPERTIES COMPILE_DEFINITIONS
    DIR2="/home/xxx/c.i")
set_source_files_properties (a.cpp PROPERTIES COMPILE_DEFINITIONS
    DIR3="/home/xxx/d.i")

结果:只有最后定义DIR3可以在make编译时在a.cpp中识别,前两个在make阶段报告未定义.

有什么建议？

谢谢!