Re: Removing unneeded self joins - Mailing list pgsql-hackers

From Andrei Lepikhov
Subject Re: Removing unneeded self joins
Date
Msg-id 0c5f461f-e26f-4fd9-9700-91b0e9c38fd6@postgrespro.ru
Whole thread Raw
In response to Re: Removing unneeded self joins  (Alexander Lakhin <exclusion@gmail.com>)
Responses Re: Removing unneeded self joins
List pgsql-hackers
On 29/12/2023 12:00, Alexander Lakhin wrote:
> Hi Alexander,
> 
> 23.10.2023 14:29, Alexander Korotkov wrote:
>> Fixed all of the above. Thank you for catching this!
> 
> I've discovered that starting from d3d55ce57 the following query:
> CREATE TABLE t(a int PRIMARY KEY);
> 
> WITH tt AS (SELECT * FROM t)
> UPDATE t SET a = tt.a + 1 FROM tt
> WHERE tt.a = t.a RETURNING t.a;
> 
> triggers an error "variable not found in subplan target lists".
> (Commits 8a8ed916f and b5fb6736e don't fix this, unfortunately.)

Thanks for the report!
The problem is with the resultRelation field. We forget to replace the 
relid here.
Could you check your issue with the patch in the attachment? Does it 
resolve this case?

-- 
regards,
Andrei Lepikhov
Postgres Professional

Attachment

pgsql-hackers by date:

Previous
From: David Rowley
Date:
Subject: Re: Revise the Asserts added to bimapset manipulation functions
Next
From: "Andrey M. Borodin"
Date:
Subject: Re: Transaction timeout