| From: | Pavel Stehule <pavel(dot)stehule(at)gmail(dot)com> | 
|---|---|
| To: | Robert Haas <robertmhaas(at)gmail(dot)com> | 
| Cc: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, Heikki Linnakangas <heikki(dot)linnakangas(at)enterprisedb(dot)com>, Teodor Sigaev <teodor(at)sigaev(dot)ru>, PostgreSQLHackers <pgsql-hackers(at)postgresql(dot)org> | 
| Subject: | Re: patch: tsearch - some memory diet | 
| Date: | 2010-10-04 06:05:48 | 
| Message-ID: | AANLkTik7YUnqM-7KHOYyRSw+Q4sFvKLvNjaTjfVFnFz5@mail.gmail.com | 
| Views: | Whole Thread | Raw Message | Download mbox | Resend email | 
| Thread: | |
| Lists: | pgsql-hackers | 
Hello
2010/10/4 Robert Haas <robertmhaas(at)gmail(dot)com>:
> On Oct 3, 2010, at 7:02 PM, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
>> It's not at all apparent that the code is even
>> safe as-is, because it's depending on the unstated assumption that that
>> static variable will get reset once per dictionary.  The documentation
>> is horrible: it doesn't really explain what the patch is doing, and what
>> it does say is wrong.
>
> Yep. We certainly would need to convince ourselves that this is correct before applying it, and that is all kinds of non-obvious.
>
what is good documentation?
This patch doesn't do more, than it removes palloc overhead on just
one structure of TSearch2 ispell dictionary. It isn't related to some
static variable - the most important is fact so this memory is
unallocated by dropping of memory context.
Regards
Pavel Stehule
> ...Robert
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Hitoshi Harada | 2010-10-04 06:36:48 | Re: wip: functions median and percentile | 
| Previous Message | Itagaki Takahiro | 2010-10-04 03:45:02 | Re: I: About "Our CLUSTER implementation is pessimal" patch |