memcpy vs slice copy

[bearophile]

For reference here's a simple C version:

#include "stdlib.h"
#include "string.h"
#include "stdio.h"

#define N 100000000
#define L 6

char h[L] = "hello\n";

int main() {
    char *ptr;
    if (N <= 100)
        ptr = malloc(N * L + 1); // the +1 is for the final printing
    else
        ptr = malloc(N * L);

    char *p = ptr;
    int i;

    for (i = 0; i < N; i++) {
        memcpy(p, h, L);
        p += L;
    }

    // to see if it works decrease N
    if (N <= 100) {
        ptr[N * L] = '\0';
        printf("%s", ptr);
    }
    return 0;
}

It takes 0.59 s to run on the same PC and OS of mine, both using GCC and LLVM-GCC.
(Note that the loop in the code is a bit different now, the end result is the same).

The ASM of the inner loop:

L:  movl    _h, %eax
    movl    %eax, (%edx)
    movzwl  _h+4, %eax
    movw    %ax, 4(%edx)
    addl    $6, %edx
    cmpl    %ecx, %edx
    jne L

Bye,
bearophile