From: | Petr Jelinek <petr(at)2ndquadrant(dot)com> |
---|---|
To: | Michael Paquier <michael(dot)paquier(at)gmail(dot)com>, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
Cc: | PostgreSQL mailing lists <pgsql-bugs(at)postgresql(dot)org> |
Subject: | Re: Re: sampling.c and potential divisions by 0 ang log(0) with tablesample and ANALYZE in 9.5 |
Date: | 2015-07-02 02:27:39 |
Message-ID: | 5594A19B.9090503@2ndquadrant.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-bugs |
On 2015-07-02 02:51, Michael Paquier wrote:
> On Wed, Jul 1, 2015 at 3:52 AM, Tom Lane wrote:
>> Rather than expanding the API of port.h still further
>> [...]
>
> The current trend, as far as I am understanding it, is not to increase
> the amount of code in src/port, but actually to reduce it and to rely
> a maximum on what modern platforms offer. See for example the recent
> discussion about rint and numeric rounding here:
> http://www.postgresql.org/message-id/flat/20150320194337(dot)2573(dot)72944(at)wrigleys(dot)postgresql(dot)org#20150320194337(dot)2573(dot)72944@wrigleys.postgresql.org
> We are keeping rint around because MSVC just added it recently in
> MS2013 even if the POSIX spec has it for a long time, but I am sure
> that we would drop it asap from the code tree if we could.
>
Which is completely moot point here because the sampling depends on
pg_erand48 on every platform no matter what the platform provides.
--
Petr Jelinek http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
From | Date | Subject | |
---|---|---|---|
Next Message | joao.antunes | 2015-07-02 10:44:59 | BUG #13480: RE: #11732 and also possibly #12538 - Possible serializable isolation lepessimistic locking problem |
Previous Message | Michael Paquier | 2015-07-02 00:51:30 | Re: Re: sampling.c and potential divisions by 0 ang log(0) with tablesample and ANALYZE in 9.5 |