| From: | John Naylor <johncnaylorls(at)gmail(dot)com> |
|---|---|
| To: | Zsolt Parragi <zsolt(dot)parragi(at)percona(dot)com> |
| Cc: | Aleksander Alekseev <aleksander(at)tigerdata(dot)com>, pgsql-hackers(at)lists(dot)postgresql(dot)org |
| Subject: | Re: [PATCH] Refactor *_abbrev_convert() functions |
| Date: | 2026-09-23 08:59:56 |
| Message-ID: | CANWCAZYeZmvovQNNXPfQsp6P=p4__MWEHrK3Yp5f0E=c_qb+3w@mail.gmail.com |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
On Wed, Sep 23, 2026 at 3:50 PM Zsolt Parragi <zsolt(dot)parragi(at)percona(dot)com> wrote:
>
> Hello
>
> - h = DatumGetUInt32(hash_uint32(k->dboid));
> - h ^= DatumGetUInt32(hash_any((const unsigned char *) k->channel,
> - strnlen(k->channel, NAMEDATALEN)));
> + h = murmurhash32(k->dboid);
> + h ^= hash_bytes((const unsigned char *) k->channel,
> + strnlen(k->channel, NAMEDATALEN));
>
> I see that this change was discussed previously, but shouldn't it be
> at least mentioned in the commit message?
I'm thinking now to leave the change to hash_uint32 out now, since
there are places elsewhere that do round-trip casting via hash_uint32,
and that could be material a separate patch, leaving this patch only
touching hash_any. I'll make the change my self and commit soon.
--
John Naylor
Amazon Web Services
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Peter Eisentraut | 2026-09-23 09:00:28 | Re: Fix conversion warnings in headers |
| Previous Message | Karina Litskevich | 2026-09-23 08:57:47 | Re: Vacuum statistics |