You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Is it possible to change the length types to uint64_t so that the algorithms work for very long arrays as well (~2^35 of 8-byte words)?
I found another library https://github.com/kimwalisch/libpopcnt, where they use uint64_t, but unfortunately, it doesn't have your nice functions for computing intersection and union.
The text was updated successfully, but these errors were encountered:
Although the function interface is f(const uint8_t* data, size_t size) you are free to pass any primitive type as input by casting. For example, for 64-bit values you can call it like this: f((const uint64_t*)data, size*sizeof(uint64_t)). Internally, all computations are performed on 64-bit words.
Importantly, this is true for all subroutines in this library with the exception of the pospopcnt family of operations which are limited to 16-bit input words.
Dear Marcus,
Is it possible to change the length types to
uint64_t
so that the algorithms work for very long arrays as well (~2^35 of 8-byte words)?I found another library https://github.com/kimwalisch/libpopcnt, where they use uint64_t, but unfortunately, it doesn't have your nice functions for computing intersection and union.
The text was updated successfully, but these errors were encountered: