From: | Bruce Momjian <maillist(at)candle(dot)pha(dot)pa(dot)us> |
---|---|
To: | vadim(at)sable(dot)krasnoyarsk(dot)su (Vadim B(dot) Mikheev) |
Cc: | hackers(at)postgreSQL(dot)org (PostgreSQL-development) |
Subject: | Re: exec_tlist_length |
Date: | 1998-07-20 20:52:17 |
Message-ID: | 199807202052.QAA03150@candle.pha.pa.us |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
> Vadim (or anyone else), can you comment on the use of
> exec_tlist_length() vs. ExecTargetListLength(). This was changed in
> Mariposa by removing the first one.
>
> The first is called only in the planner.c, and computes the length as:
>
> len = 0;
> foreach(tl, targetlist)
> {
> curTle = lfirst(tl);
>
> if (curTle->resdom != NULL)
> len++;
> }
> return len;
>
> while ExecTargetListLength() uses:
>
> len = 0;
> foreach(tl, targetlist)
> {
> curTle = lfirst(tl);
>
> if (curTle->resdom != NULL)
> len++;
> else
> len += curTle->fjoin->fj_nNodes;
> }
> return len;
>
> The second counts resdom as one, or add fj_nNodes. Which is correct, or
> are they used for different purposes. Seems like the second is more
> correct.
I have done some more research and found fj_nNodes is not used
(referenced in the FIXED_SETS code), so I have removed the first
function and make them all reference the second.
--
Bruce Momjian | 830 Blythe Avenue
maillist(at)candle(dot)pha(dot)pa(dot)us | Drexel Hill, Pennsylvania 19026
+ If your life is a hard drive, | (610) 353-9879(w)
+ Christ can be your backup. | (610) 853-3000(h)
From | Date | Subject | |
---|---|---|---|
Next Message | Keith Parks | 1998-07-20 21:21:29 | Re: [HACKERS] s_lock.h problem on S/Linux |
Previous Message | Bruce Momjian | 1998-07-20 20:45:24 | cvs and empty directories |