Content-Type: text/plain;
	charset=utf-8
Mime-Version: 1.0 (Mac OS X Mail 16.0 \(3864.400.21\))
Subject: Re: Fix memory leak in postmasterMain
From: Chao Li <li.evan.chao@gmail.com>
In-Reply-To: <mwikkx643zp3f5qmx27crf2d2ngwgupc4om4kfq4gyc6vukjom@kyimpnconhdv>
Date: Thu, 23 Apr 2026 13:22:04 +0800
Cc: Fujii Masao <masao.fujii@gmail.com>,
 Henrik TJ <henrik@0x48.dk>,
 PostgreSQL Hackers <pgsql-hackers@postgresql.org>
Content-Transfer-Encoding: quoted-printable
Message-Id: <E07322B5-FDE6-4334-BBE5-01D5E5208C61@gmail.com>
References: <e84ede22-1b46-0780-64e4-5db4a1e5236f@0x48.dk>
 <3e890874-8204-e1c2-8def-3a02cbad2232@0x48.dk>
 <1658D50F-A8B3-406A-94ED-FA8C7F9DA90B@gmail.com>
 <CAHGQGwEn3oEad-wR1XoEbe1R-pB1EOycRKjEj+5Pp38mjgiF+Q@mail.gmail.com>
 <mwikkx643zp3f5qmx27crf2d2ngwgupc4om4kfq4gyc6vukjom@kyimpnconhdv>
To: Andres Freund <andres@anarazel.de>
Archived-At: 
 <https://www.postgresql.org/message-id/E07322B5-FDE6-4334-BBE5-01D5E5208C61%40gmail.com>
Precedence: bulk


> On Apr 23, 2026, at 01:55, Andres Freund <andres@anarazel.de> wrote:
>=20
> Hi,
>=20
> On 2026-04-23 00:29:29 +0900, Fujii Masao wrote:
>> On Wed, Apr 22, 2026 at 3:46=E2=80=AFPM Chao Li =
<li.evan.chao@gmail.com> wrote:
>>>=20
>>> =46rom my experience, most of the time the postmaster is started =
with -D. On Linux and macOS, that path can be quite long, PATH_MAX is =
often 4096 on many Unix-like systems, and I am not sure about Windows. =
So I think this leak is worth fixing.
>>=20
>> We can also free()/pfree() userDoption in postgres.c and bootstrap.c?
>>=20
>> Since userDoption typically uses only a small amount of memory, I'm =
not sure
>> this patch provides much practical benefit from a memory-leak =
perspective.
>=20
> I don't think this is a leak at all. We *never* can reach the end of =
the scope
> in which userDoption is allocated. We abort() if the end of =
PostmasterMain()
> is ever reached.  What is the leak here supposed to actually be?
>=20

Yes, calling it =E2=80=9Cmemory leak=E2=80=9D is too strict. Usually, =
memory leak implies repeatedly losing memory over time, but in this =
case, userDoption is no longer used after feeding to =
SelectConfigFiles(). So calling it =E2=80=9Cunnecessary retained =
memory=E2=80=9D might be more accurate.

In theory, it occupies at most PATH_MAX bytes, but in practice it should =
be much less than that. So, I agree the benefit of fixing is trivial.

>=20
> ISTM those strdup()s should actually be pstrdup()s?  I suspect =
changing that
> would also silence valgrind.
>=20

I also had the question when I read the code. But looks like startup =
phase all uses malloc/strdup etc. C functions, for example =
SelectConfigFiles() in guc.c also uses malloc/free. I am not sure what =
is the standard.

Best regards,
--
Chao Li (Evan)
HighGo Software Co., Ltd.
https://www.highgo.com/