How can I create CUDAKernel objects from compiled PTX code that implements dynamic parallelism?
8 views (last 30 days)
I have a ".cu" file that uses dynamic parallelism - meaning it uses a CUDA kernel to launch another CUDA kernel (all on the GPU). The code compiles fine as a PTX file. But when I go to create the CUDAKernel object in MATLAB, I get the following error:
>> parent = parallel.gpu.CUDAKernel('dp_test.ptx','dp_test.cu','test_parent');
Error using parallel.gpu.CUDAKernel
An error occurred during PTX compilation of <image>.
The information log was:
The error log was:
ptxas : fatal error : Unresolved extern function 'cudaGetParameterBuffer
The CUDA error code was: CUDA_ERROR_NO_BINARY_FOR_GPU.
This question is similar to one posted on StackOverflow (not by me): http://stackoverflow.com/questions/21175879/how-to-compile-and-launch-cuda-codes-using-dynamic-parallelism-from-matlab
I appreciate any and all help. Thanks!