I have two questions.
Do realloc()
and memcpy()
copy the entries in an array to another in a way faster than just iterating on each element O(N)
? If the answer is yes then what do you think is its complexity ?
If the size allocated is smaller than the original size, does realloc()
copy the entries to somewhere else or just leave them as they are decreasing the size of the array ?
1 - No. They copy a block at a time. See http://www.embedded.com/design/configurable-systems/4024961/Optimizing-Memcpy-improves-speed for a pretty good analysis.
2 - This is implementation dependent. See http://www.gnu.org/software/libtool/manual/libc/Changing-Block-Size.html for glibc details. "In several allocation implementations, making a block smaller sometimes necessitates copying it"