From: | Hitoshi Harada <umi(dot)tanuki(at)gmail(dot)com> |
---|---|
To: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
Cc: | Robert Haas <robertmhaas(at)gmail(dot)com>, pgsql-committers(at)postgresql(dot)org |
Subject: | Re: pgsql: Support RIGHT and FULL OUTER JOIN in hash joins. |
Date: | 2011-05-25 02:01:46 |
Message-ID: | BANLkTi=vAGgYAS21zu0kv1SsfmvTkDbRMg@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-committers |
2011/5/24 Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>:
> Hitoshi Harada <umi(dot)tanuki(at)gmail(dot)com> writes:
>> It looks to me like mergejoin_allowed should be initialized to false.
>> If enable_mergejoin is off and jointype != JOIN_FULL then mergejoin is
>> not allowed, isn't it? Sorry for noise if it's only my confusing.
>
> No, the code is correct as-is: we need the flag to be set true by
> default, because the way that the function is using it, we'll also skip
> some nestloop possibilities when it's false. Note the comment for step
> 2.
>
> It could be that a different name for that flag variable would be a
> better idea, but neither Robert nor I could come up with a better one.
Ah, ok. I think I now understand it. "mergejoin_doable" flag. Thanks.
Regards,
--
Hitoshi Harada
From | Date | Subject | |
---|---|---|---|
Next Message | Bruce Momjian | 2011-05-25 04:00:17 | pgsql: Add C comment about why we don't spell out "month" in interval v |
Previous Message | Tom Lane | 2011-05-24 22:52:22 | pgsql: Grammar cleanup for src/test/isolation/README |