Hi,
I have a discussion with Teodor over the phone and we agree
this is the best for 7.2.*. Thanks Tom for the help.
btw, I think it should be noticed somewhere in documentation for developers
that pointers to "int" and "long" which are the same on 32-bit machine,
are different on the 64-bit machine.
Oleg
PS. For more than year of our GiST development we got the first report
from 64-bit machine and I think it's a good sign. This year we must
add concurrency support. We already had discussion with
Joseph Hellerstein ( the 'father' of the GiST ) about concurrency support
and perhaps we'll go along the paper by Marcel Kornacker.
On Fri, 8 Feb 2002, Tom Lane wrote:
> Actually, there is a third possibility, which would fix the problem
> without requiring any changes in the picksplit functions. You could
> do this:
>
> char *storage;
>
> storage = palloc(MAXALIGN(VARHDRSZ) + (*len + 1) * sizeof(GISTENTRY));
> entryvec = (bytea *) (storage + MAXALIGN(VARHDRSZ) - VARHDRSZ);
>
> use entryvec as before, except final pfree is pfree(storage)
>
> Grotty as heck, but probably the right answer for 7.2.1 to avoid the
> initdb issues.
>
> For 7.3 we could do it the other, cleaner way.
>
> regards, tom lane
>
> ---------------------------(end of broadcast)---------------------------
> TIP 1: subscribe and unsubscribe commands go to majordomo@postgresql.org
>
Regards, Oleg
_____________________________________________________________
Oleg Bartunov, sci.researcher, hostmaster of AstroNet,
Sternberg Astronomical Institute, Moscow University (Russia)
Internet: oleg@sai.msu.su, http://www.sai.msu.su/~megera/
phone: +007(095)939-16-83, +007(095)939-23-83
