Re: Memory leak on subquery as scalar operand - Mailing list pgsql-bugs

From David Rowley
Subject Re: Memory leak on subquery as scalar operand
Date
Msg-id CAApHDvrSbQkPUacnaonQzw1efFJM6HE5we-i9VVogBfEKGCCiA@mail.gmail.com
Whole thread Raw
In response to Memory leak on subquery as scalar operand  (Lauri Laanmets <pcspets@gmail.com>)
Responses Re: Memory leak on subquery as scalar operand  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-bugs
On Mon, 31 Oct 2022 at 23:03, Lauri Laanmets <pcspets@gmail.com> wrote:
> There seems to be a memory leak in PostgreSQL for Linux (doesn't happen on Windows). At least in versions 12 to 14
(andnot in 9.5). It can be reproduced with the following SQL script on an empty database.
 
>
> * When repeated, Linux process starts using more and more RAM until OOM-killer will kill it.
> * 'pg_backend_memory_contexts' doesn't show any increase.
> * (gdb) p MemoryContextStats(TopMemoryContext) - doesn't show any increase.

Does it still OOM if you do:

SET jit TO off; ?

David



pgsql-bugs by date:

Previous
From: David Rowley
Date:
Subject: Re: Strange sort node/explain result
Next
From: Tom Lane
Date:
Subject: Re: Memory leak on subquery as scalar operand