maria-developers team mailing list archive

[Sergey Petrunia <sergey@xxxxxxxxxxx>]

Hi Rex,

I am looking at the latest commit:

commit 9eba044b9710bbdd8195e887cfdd4d62233942ca (HEAD -> bb-MDEV-21092-21095-29997-optimizer-trace-updates,
Author: Rex <rex.johnston@xxxxxxxxxxx>
Date:   Fri Dec 2 16:46:54 2022 +1200

MDEV-21092,MDEV-21095,MDEV-29997: Optimizer Trace for index condition pushdown, partition pruning, exists-to-in

Add Optimizer Tracing for:
- Index Condition Pushdown
- Partition Pruning
- Exists-to-IN optimization

and trying this testcase:
== testcase ==

--source include/have_sequence.inc

create table t1 (
  a int,
  b int,
  key(a)
);

insert into t1 select seq, seq from seq_1_to_10000;

--optimizer_trace
explain format=json select * from t1 where a < 20 and (a+1<a+2) and b=3333;
drop table t1;

== Testcase ends ==

In the output I see:

== Output ==

          {
            "make_join_readinfo": [
              {
                "table": "t1",
                "index_condition": "t1.b = 3333" // WRONG-1
              }
            ]
          }
        ]
      }
    },
    {
      "attaching_conditions_to_tables": {
        "attached_conditions_computation": [],
        "attached_conditions_summary": [
          {
            "table": "t1",
            "attached_condition": "t1.b = 3333",
            "index_condition": "t1.a < 20 and t1.a + 1 < t1.a + 2"
          }
        ]
      }
    },

== Output ends ==

note the line marked "WRONG-1". 

I also think some of the recent changes in the patch go in the wrong
direction. Let me take over, I think I'll make a patch to fix both.

BR
 Sergei
-- 
Sergei Petrunia, Software Developer
MariaDB Corporation | Skype: sergefp | Blog: http://petrunia.net