Re: [PATCH] Small optimization across postgres (remove strlenduplicate usage)

On Mon, 20 Apr 2020 at 11:24, Ranier Vilela <ranier.vf@gmail.com> wrote:
> I tried: https://godbolt.org with:
>
> -O2:
>
> f1:
> int main (int argv, char **argc)
> {
>     return strlen(argc[0]) == 0;
> }
>
> f1: Assembly
> main:                                   # @main
>         mov     rcx, qword ptr [rsi]
>         xor     eax, eax
>         cmp     byte ptr [rcx], 0
>         sete    al
>         ret
>
> f2:
> int main (int argv, char **argc)
> {
>     return argc[0] == '\0';
> }
>
> f2: Assembly
>
> main:                                   # @main
>         xor     eax, eax
>         cmp     qword ptr [rsi], 0
>         sete    al
>         ret
>
> For me clearly str [0] == '\ 0', wins.

I think you'd want to use argc[0][0] == '\0' or *argc[0] == '\0'.
Otherwise you appear just to be checking if the first element in the
argc pointer array is set to NULL, which is certainly not the same as
an empty string.