Horton

Re: [PATCH] Incremental sort (was: PoC: Partial sort)

James Coleman <jtc331@gmail.com>

View thread View on postgr.es
From: James Coleman <jtc331@gmail.com>
To: Andreas Karlsson <andreas@proxel.se>
Cc: Tomas Vondra <tomas.vondra@2ndquadrant.com>, Michael Paquier <michael@paquier.xyz>, Rafia Sabih <rafia.pghackers@gmail.com>, Peter Geoghegan <pg@bowt.ie>, Simon Riggs <simon@2ndquadrant.com>, Shaun Thomas <shaun.thomas@2ndquadrant.com>, Dmitry Dolgov <9erthalion6@gmail.com>, Alexander Korotkov <a.korotkov@postgrespro.ru>, Tom Lane <tgl@sss.pgh.pa.us>, PostgreSQL Developers <pgsql-hackers@lists.postgresql.org>, Alvaro Herrera <alvherre@2ndquadrant.com>
Date: 2020-03-23T00:33:46Z
Lists: pgsql-hackers 
On Sun, Mar 22, 2020 at 6:02 PM Andreas Karlsson <andreas@proxel.se> wrote:
>
> On 3/21/20 1:56 AM, Tomas Vondra wrote:
> > I've looked at v38 but it seems it's a bit broken by some recent explain
> > changes (mostly missing type in declarations). Attached is v39 fixing
> > those issues, and including a bunch of fixes based on a review - most of
> > the changes is in comments, so I've instead kept them in separate "fix"
> > patches after each part.
> >
> > In general I'm mostly happy with the current shape of the patch, and
> > unless there are some objections I'd like to get some of it committed
> > sometime next week.
>
> Hi,
>
> I haven't had any time to look at the patch yet but when compiling it
> and running the tests GCC gave me a warning. The tests passed btw.
>
> gcc (Debian 8.3.0-6) 8.3.0
>
> explain.c: In function ‘show_incremental_sort_group_info’:
> explain.c:2772:39: warning: passing argument 2 of ‘lappend’ discards
> ‘const’ qualifier from pointer target type [-Wdiscarded-qualifiers]
>      methodNames = lappend(methodNames, sortMethodName);
>                                         ^~~~~~~~~~~~~~
> In file included from ../../../src/include/access/xact.h:20,
>                   from explain.c:16:
> ../../../src/include/nodes/pg_list.h:509:14: note: expected ‘void *’ but
> argument is of type ‘const char *’
>   extern List *lappend(List *list, void *datum);
>                ^~~~~~~
> explain.c:2772:39: warning: passing 'const char *' to parameter of type
> 'void *' discards qualifiers
>        [-Wincompatible-pointer-types-discards-qualifiers]
>                          methodNames = lappend(methodNames, sortMethodName);
>                                                             ^~~~~~~~~~~~~~
> ../../../src/include/nodes/pg_list.h:509:40: note: passing argument to
> parameter 'datum' here
> extern List *lappend(List *list, void *datum);

So if we naively get rid of the const on the variable declaration in
question, then we get this warning instead:

explain.c: In function ‘show_incremental_sort_group_info’:
explain.c:2770:27: warning: initialization discards ‘const’ qualifier
from pointer target type [-Wdiscarded-qualifiers]
    char *sortMethodName = tuplesort_method_name(methodCell->int_value);

So on the face of it we have a bit of a no-win situation. The function
tuple_sort_method_name returns a const, but lappend wants a non-const.
I'm not sure what the project style preference is here: we could cast
the result as (char *) to drop the const qualifier, but that's frowned
upon some places. Alternatively we could make a new non-const copy of
string. Which is preferable in the postgres project style?

James

Commits

  1. Further adjustments to Hashagg EXPLAIN ANALYZE output

  2. Rework EXPLAIN format for incremental sort

  3. Fix typos and improve incremental sort comments

  4. Stabilize incremental_sort tests

  5. Minor improvements in Incremental Sort explain

  6. Consider Incremental Sort paths at additional places

  7. Fix representation of SORT_TYPE_STILL_IN_PROGRESS.

  8. Fix failures in incremental_sort due to number of workers

  9. Fix show_incremental_sort_info with force_parallel_mode

  10. Implement Incremental Sort

  11. Fix handling of "Subplans Removed" field in EXPLAIN output.

  12. Fix EXPLAIN (SETTINGS) to follow policy about when to print empty fields.

  13. Ensure plpgsql result tuples have the right composite type marking.

  14. Propagate sort instrumentation from workers back to leader.

  15. Make new regression test case parallel-safe, and improve its output.

  16. Push limit through subqueries to underlying sort, where possible.

  17. Fix inappropriate printing of never-measured times in EXPLAIN.

  18. Fix some infelicities in EXPLAIN output for parallel query plans.