0

I'm using Visual Studio 2015 in combination with CUDA toolkit 8.0, and would like to make use of the dynamic parallelism features. I've found lots and lots of blog posts about how to use the feature, but none seem to concern themselves with what requirements exist in terms of compiler flags in order to get the program to compile in the first place. I have as of yet failed to find any information on this through google searches.

My current efforts have stranded on fiddling with various settings in the CUDA C/C++ section in the project properties, but anything I change appears to produce seemingly random and arbitrary linker errors. Googling on these yields some SO questions, amongst others, which do not seem to agree on what switches are truly necessary.

Can someone perhaps shed some light on this?

Bartvbl
  • 2,878
  • 3
  • 28
  • 45
  • 1
    http://stackoverflow.com/questions/19287461/compiling-code-containing-dynamic-parallelism-fails – talonmies Feb 15 '17 at 13:05
  • @talonmies Once you actually go and ask about this stuff you figure out you were not linking the cuda device runtime correctly. My project finally compiles. – Bartvbl Feb 15 '17 at 13:11

0 Answers0