Re: An out-of-date comment in nodeIndexonlyscan.c - Mailing list pgsql-hackers

From David Rowley
Subject Re: An out-of-date comment in nodeIndexonlyscan.c
Date
Msg-id CAApHDvog5unm6LQ-68B4gyhYF-Gf+6-87fns1zTFvwbrFv0wwg@mail.gmail.com
Whole thread Raw
In response to Re: An out-of-date comment in nodeIndexonlyscan.c  (Thomas Munro <thomas.munro@gmail.com>)
Responses Re: An out-of-date comment in nodeIndexonlyscan.c
List pgsql-hackers
On Mon, 14 Jun 2021 at 00:02, Thomas Munro <thomas.munro@gmail.com> wrote:
> Here's a highly experimental patch I came up with that seems to
> produce the right results in simple cases, without (yet) involving the
> planner.

+ /* Find all equality quals. */
+ for (int i = 0; i < n_scan_keys; ++i)
+ {
+ if (scan_keys[i].sk_strategy == BTEqualStrategyNumber)
+ attnos[nattnos++] = scan_keys[i].sk_attno;
+ }
+
+ /* Are all attributes covered? */
+ /* XXX is this check enough or do we need to work harder? */
+ qsort(attnos, nattnos, sizeof(AttrNumber), compare_int16);
+ nattnos = qunique(attnos, nattnos, sizeof(AttrNumber), compare_int16);
+ if (nattnos == index->rd_index->indnkeyatts)

I think a more optimal and nicer way of doing that would be setting
bits in a Bitmapset then checking bms_num_members is equal to
n_scan_keys.

David



pgsql-hackers by date:

Previous
From: Ranier Vilela
Date:
Subject: Re: Signed vs Unsigned (take 2) (src/backend/storage/ipc/procarray.c)
Next
From: Tomas Vondra
Date:
Subject: Re: a path towards replacing GEQO with something better