| From: | Andres Freund <andres(at)anarazel(dot)de> |
|---|---|
| To: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
| Cc: | Mark Dilger <mark(dot)dilger(at)enterprisedb(dot)com>, Arjan van de Ven <arjan(at)linux(dot)intel(dot)com>, pgsql-hackers(at)postgresql(dot)org |
| Subject: | Re: src/port/snprintf.c: Optimize the common base=10 case in fmtint |
| Date: | 2021-10-26 18:15:35 |
| Message-ID: | 20211026181535.6cncdjfpr4u3frr4@alap3.anarazel.de |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Hi,
On 2021-10-26 13:51:55 -0400, Tom Lane wrote:
> Mark Dilger <mark(dot)dilger(at)enterprisedb(dot)com> writes:
> > It appears fmtint only has three options for base, being 10, 16, and 8. Have you profiled with either of the others special cased as well? I don't see much use in optimizing for octal, but hexadecimal is used quite a bit in wal with patterns like "%08X%08X%08X".
>
> I'd be inclined to just hard-wire the three allowed cases, and not have
> an arbitrary-divisor code path at all.
Yea, I came to the same conclusion. But I'd implement it by moving the
division into a separate inline function called from the switch. I tested that
locally and it works, but I got sidetracked by [1].
Greetings,
Andres Freund
[1] https://postgr.es/m/20211026180454.xcjmu3kwmn3tka57%40alap3.anarazel.de
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Andres Freund | 2021-10-26 18:31:40 | Re: changes in pgport etc doesn't cause client programs to be relinked |
| Previous Message | Arjan van de Ven | 2021-10-26 18:13:14 | Re: src/port/snprintf.c: Optimize the common base=10 case in fmtint |