Re: GiST penalty functions [PoC]

Поиск
Список
Период
Сортировка
От Greg Stark
Тема Re: GiST penalty functions [PoC]
Дата
Msg-id CAM-w4HNf34gYFKug2Q2gnbpA8hDvtTbhrNoqMHgYD0za=M0yPg@mail.gmail.com
обсуждение исходный текст
Ответ на Re: GiST penalty functions [PoC]  (Andrew Borodin <borodin@octonica.com>)
Ответы Re: GiST penalty functions [PoC]  (Andrew Borodin <borodin@octonica.com>)
Re: GiST penalty functions [PoC]  (Andrew Borodin <borodin@octonica.com>)
Список pgsql-hackers
On Thu, Sep 8, 2016 at 9:29 AM, Andrew Borodin <borodin@octonica.com> wrote:
>>autoconf check for IEEE 754 floats
> Autoconf man says folowing:
>>it is safe to assume IEEE-754 in most portable code these days
> https://www.gnu.org/software/autoconf/manual/autoconf.html#Floating-Point-Portability


Personally I wouldn't be very happy about an IEEE754 assumption. I did
go to the trouble of testing Postgres on a VAX and we fixed the few
instances where it had such dependencies for a net gain. If we
intentionally put a dependency in in one place then we'll never be
able to determine anywhere else where there are unintentional
dependencies.

I haven't followed the thread closely but I'm puzzled why you would
need to use bit twiddling to set a floating point number. Isn't there
a perfectly good way to calculate the value you want using ldexp() and
other standard C library functions?

-- 
greg



В списке pgsql-hackers по дате отправления:

Предыдущее
От: Tom Lane
Дата:
Сообщение: Re: Re: GiST optimizing memmoves in gistplacetopage for fixed-size updates [PoC]
Следующее
От: Adam Brightwell
Дата:
Сообщение: Re: COPY command with RLS bug