From: Tom Lane <tgl@sss.pgh.pa.us>
To: Matthew Clark <mclark@drmatthewclark.com>
cc: "pgsql-bugs@lists.postgresql.org" <pgsql-bugs@lists.postgresql.org>
Subject: Re: server crash on raspberry pi for large queries
In-reply-to: <2123352386.9759142.1724114730509@mail.yahoo.com>
References: <2123352386.9759142.1724114730509.ref@mail.yahoo.com>
 <2123352386.9759142.1724114730509@mail.yahoo.com>
Comments: In-reply-to Matthew Clark <mclark@drmatthewclark.com>
	message dated "Tue, 20 Aug 2024 00:45:30 -0000"
MIME-Version: 1.0
Content-Type: text/plain; charset="UTF-8"
Content-ID: <1527795.1724159772.1@sss.pgh.pa.us>
Content-Transfer-Encoding: quoted-printable
Date: Tue, 20 Aug 2024 09:16:12 -0400
Message-ID: <1527796.1724159772@sss.pgh.pa.us>
Archived-At: 
 <https://www.postgresql.org/message-id/1527796.1724159772%40sss.pgh.pa.us>
Precedence: bulk

Matthew Clark <mclark@drmatthewclark.com> writes:
> issue: when issuing a large query ,=C2=A0 e.g.=C2=A0=C2=A0 "Select count=
(*) from temperature" on a table with > 150M rows the system crashes.
> =C2=A0 example error:=C2=A0=C2=A0 LOG:=C2=A0 server process (PID 20037) =
was terminated by signal 4: Illegal instruction=C2=A0 no more diagnostic i=
nformation is generated.
> Workaround : this stops happening when jit_above_cost is set to -1 in po=
stgresql.conf=C2=A0=C2=A0 After changing this setting the problem stops.

What this sounds like is a memory leak in the JIT stuff.  We fixed
one such issue last year, but perhaps there's more.  Can you provide
a self-contained test case?  Also, please be more specific about
which Linux version you are using, and which LLVM version.

			regards, tom lane