From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
---|---|
To: | Alvaro Herrera <alvherre(at)commandprompt(dot)com> |
Cc: | richhguard-monotone <richhguard-monotone(at)yahoo(dot)co(dot)uk>, pgsql-hackers <pgsql-hackers(at)postgresql(dot)org> |
Subject: | Re: PATCH: CreateComments: use explicit indexing for ``values'' |
Date: | 2011-06-14 14:30:28 |
Message-ID: | 27858.1308061828@sss.pgh.pa.us |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
Alvaro Herrera <alvherre(at)commandprompt(dot)com> writes:
> Excerpts from richhguard-monotone's message of lun jun 13 16:10:17 -0400 2011:
>> Do you have any advice of how to handle the inner loops, such as those initializing ``stakindN''. The entries before can be handled just like in this patch, by using the symbolic constants.
> Based on Tom's comments, I'd submit the patch without that bit, at least
> as a first step.
He already did no?
I did think of a possible way to rewrite update_attstats: instead of
for (k = 0; k < STATISTIC_NUM_SLOTS; k++)
{
values[i++] = ObjectIdGetDatum(stats->staop[k]); /* staopN */
}
do
for (k = 0; k < STATISTIC_NUM_SLOTS; k++)
{
values[Anum_pg_statistic_staop1 - 1 + k] = ObjectIdGetDatum(stats->staop[k]);
}
etc. However, it's not clear to me whether this is really an
improvement. Opinions?
regards, tom lane
From | Date | Subject | |
---|---|---|---|
Next Message | Alvaro Herrera | 2011-06-14 14:35:53 | Re: ITYM DROP TABLE |
Previous Message | Cédric Villemain | 2011-06-14 14:29:36 | [WIP] cache estimates, cache access cost |