Re: Potential memory leak in contrib/intarray's g_intbig_compress - Mailing list pgsql-hackers

From Michael Paquier
Subject Re: Potential memory leak in contrib/intarray's g_intbig_compress
Date
Msg-id ZLDjyfq9cT4+Vu0u@paquier.xyz
Whole thread Raw
In response to Re: Potential memory leak in contrib/intarray's g_intbig_compress  (Matthias van de Meent <boekewurm+postgres@gmail.com>)
Responses Re: Potential memory leak in contrib/intarray's g_intbig_compress
List pgsql-hackers
On Thu, Jul 13, 2023 at 06:28:39PM +0200, Matthias van de Meent wrote:
> There are similar pfree calls in the _int_gist.c file's g_int_compress
> function, which made me think we do need to clean up after use, but
> indeed these pfrees are useless (or even harmful if bug #17888 can be
> trusted)

Indeed, all these are in a GiST temporary context.  So you'd mean
something like the attached perhaps, for both the decompress and
compress paths?
--
Michael

Attachment

pgsql-hackers by date:

Previous
From: Martin Butter
Date:
Subject: 16beta2 SQL parser: different defaults on absent_on_null
Next
From: Ashutosh Bapat
Date:
Subject: Re: logical decoding and replication of sequences, take 2