I've somehow mis-implemented atomic::increment() and atomic::decrement(). Need to bring the comments and the implementation in line with the CUDA Programming Guide description, as well as make sure that the default arguments actually do something useful (which, currently, they do not).
I've somehow mis-implemented
atomic::increment()
andatomic::decrement()
. Need to bring the comments and the implementation in line with the CUDA Programming Guide description, as well as make sure that the default arguments actually do something useful (which, currently, they do not).