It is suitable for lexicographical comparisons, or to compare nonces
and counters stored in little-endian format.
However, it is slower than sodium_memcmp().
The comparison is done in constant time for a given length.
The two numbers don't need to have the same length in bytes.
If array's length are NOT equal, the decision may be based on that, i.e. no constant time guarantee.
Preferably use this function with equal length arrays.
Comparing large numbers.
It is suitable for lexicographical comparisons, or to compare nonces and counters stored in little-endian format. However, it is slower than sodium_memcmp(). The comparison is done in constant time for a given length. The two numbers don't need to have the same length in bytes. If array's length are NOT equal, the decision may be based on that, i.e. no constant time guarantee. Preferably use this function with equal length arrays.