GUPS: Fix overflow in type cast #57
Description
#49 adds CUDA GUPS to support this blog post
In posts/gups/gups.cu there is a typecasting typo while obtaining the problem size:
- size_t n = (size_t)(1 << logn);
+ size_t n = ((size_t)1) << logn;This fix is required in order to benchmark size n = 2^31.
More fixes are required to run n = 2^32 and further, but I don't have a clue yet 😅