[Openmp-commits] [PATCH] D46185: [OpenMP] Allow nvptx sm_30 to be used as an offloading device

Greg Rodgers via Phabricator via Openmp-commits openmp-commits at lists.llvm.org
Mon May 7 09:12:34 PDT 2018


gregrodgers added a comment.

I agree that George's RMW proposed code is correct.   This was my first attempt at an RMW code.  Maybe we should implement atomicMax as a device function in architecture-specific (e.g sm_30) device library. This way the code in loop.cu can remain just a call to atomicMax.    Such an implementation would need an overloaded atomicMax.


Repository:
  rOMP OpenMP

https://reviews.llvm.org/D46185





More information about the Openmp-commits mailing list