From: Tom Lane <tgl@sss.pgh.pa.us>
To: depesz@depesz.com
cc: PostgreSQL mailing lists <pgsql-bugs@lists.postgresql.org>
Subject: Re: Warning while compiling 19beta1
In-reply-to: <3208217.1780584180@sss.pgh.pa.us>
References: <aiGNJ89PBqvq2Yyz@depesz.com> <3208217.1780584180@sss.pgh.pa.us>
Comments: In-reply-to Tom Lane <tgl@sss.pgh.pa.us>
	message dated "Thu, 04 Jun 2026 10:43:00 -0400"
MIME-Version: 1.0
Content-Type: text/plain; charset="UTF-8"
Content-ID: <3277298.1780585891.1@sss.pgh.pa.us>
Content-Transfer-Encoding: quoted-printable
Date: Thu, 04 Jun 2026 11:11:31 -0400
Message-ID: <3277299.1780585891@sss.pgh.pa.us>
Archived-At: 
 <https://www.postgresql.org/message-id/3277299.1780585891%40sss.pgh.pa.us>
Precedence: bulk

I wrote:
> hubert depesz lubaczewski <depesz@depesz.com> writes:
>> llvmjit.c: In function =E2=80=98llvm_split_symbol_name=E2=80=99:
>> llvmjit.c:1065:27: warning: assignment discards =E2=80=98const=E2=80=99=
 qualifier from pointer target type [-Wdiscarded-qualifiers]
>> 1065 |                 *funcname =3D strrchr(name, '.');
>> |                           ^

> Yeah, some of the buildfarm animals have been showing this for awhile.

Actually, after looking closer, the LLVM version isn't relevant.
What you need is a gcc new enough to know that the result of
strrchr() applied to a const char * should be considered const.

This function is ugly in some other ways too.  Will fix.

			regards, tom lane