Re: Speedup usages of pg_*toa() functions

Поиск
Список
Период
Сортировка
От Ranier Vilela
Тема Re: Speedup usages of pg_*toa() functions
Дата
Msg-id CAEudQApyRFg=anQGx6Qqt50GnDtZkU2L335140gnA=ZTb28tvA@mail.gmail.com
обсуждение исходный текст
Ответ на Re: Speedup usages of pg_*toa() functions  (David Rowley <dgrowleyml@gmail.com>)
Ответы Re: Speedup usages of pg_*toa() functions  (Andrew Gierth <andrew@tao11.riddles.org.uk>)
Список pgsql-hackers
Em ter., 9 de jun. de 2020 às 07:55, David Rowley <dgrowleyml@gmail.com> escreveu:
On Tue, 9 Jun 2020 at 22:08, Andrew Gierth <andrew@tao11.riddles.org.uk> wrote:
>
> >>>>> "David" == David Rowley <dgrowleyml@gmail.com> writes:
>
>  David> This allows us to speed up a few cases. int2vectorout() should
>  David> be faster and int8out() becomes a bit faster if we get rid of
>  David> the strdup() call and replace it with a palloc()/memcpy() call.
>
> What about removing the memcpy entirely? I don't think we save anything
> much useful here by pallocing the exact length, rather than doing what
> int4out does and palloc a fixed size and convert the int directly into
> it.

On looking back through git blame, it seems int2out and int4out have
been that way since at least 1996, before int8.c existed. int8out has
been doing it since fa838876e9f -- Include 8-byte integer type. dated
1998.  Quite likely the larger than required palloc size back then was
more of a concern. So perhaps you're right about just doing it that
way instead. With that and the ints I tested with, the int8
performance should be about aligned to int4 performance.

> For pg_ltoa, etc., I don't like adding the extra call to pg_ultoa_n - at
> least on my clang, that results in two copies of pg_ultoa_n inlined.
> How about doing it like,
>
> int
> pg_lltoa(int64 value, char *a)
> {
>     int         len = 0;
>     uint64      uvalue = value;
>
>     if (value < 0)
>     {
>         uvalue = (uint64) 0 - uvalue;
>         a[len++] = '-';
>     }
>     len += pg_ulltoa_n(uvalue, a + len);
>     a[len] = '\0';
>     return len;
> }

Written like that, wouldn't it get better?

int
pg_lltoa(int64 value, char *a)
{
    if (value < 0)
    {
        int         len = 0;
        uint64      uvalue = (uint64) 0 - uvalue;

        a[len++] = '-';
        len += pg_ulltoa_n(uvalue, a + len);
        a[len] = '\0';
        return len;
    }
else
        return pg_ulltoa_n(value, a);
}
 
regards,
Ranier Vilela

В списке pgsql-hackers по дате отправления:

Предыдущее
От: "David G. Johnston"
Дата:
Сообщение: Re: Postgres installer with pgAdmin 4.22
Следующее
От: "David G. Johnston"
Дата:
Сообщение: Re: Terminate the idle sessions