Date: Tue, 20 Aug 2024 00:45:30 +0000 (UTC)
From: Matthew Clark <mclark@drmatthewclark.com>
To: "pgsql-bugs@lists.postgresql.org" <pgsql-bugs@lists.postgresql.org>
Message-ID: <2123352386.9759142.1724114730509@mail.yahoo.com>
Subject: server crash on raspberry pi for large queries
MIME-Version: 1.0
Content-Type: multipart/alternative;
	boundary="----=_Part_9759141_267692230.1724114730508"
References: <2123352386.9759142.1724114730509.ref@mail.yahoo.com>
Content-Length: 4509
Archived-At: 
 <https://www.postgresql.org/message-id/2123352386.9759142.1724114730509%40mail.yahoo.com>
Precedence: bulk

------=_Part_9759141_267692230.1724114730508
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

postgres version:=C2=A0 PostgreSQL 15.8 (Raspbian 15.8-0+deb12u1) on arm-un=
known-linux-gnueabihf, compiled by gcc (Raspbian 12.2.0-14+rpi1)OS:=C2=A0 L=
inux pi 6.6.31+rpt-rpi-v6 #1 Raspbian 1:6.6.31-1+rpt1 (2024-05-29) armv6l G=
NU/Linux

CPU: (/proc/cpuinfo)

name=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 : ARMv6-compatible processor rev 7 (v6l)
BogoMIPS=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 : 697.95
Features=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 : half thumb fastmult vf=
p edsp java tls
CPU implementer : 0x41
CPU architecture: 7
CPU variant=C2=A0=C2=A0=C2=A0=C2=A0 : 0x0
CPU part=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 : 0xb76
CPU revision=C2=A0=C2=A0=C2=A0 : 7

Hardware=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 : BCM2835
Revision=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 : 000e
Serial=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 : 00000000d75e=
7c8b
Model=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 : Raspber=
ry Pi Model B Rev 2
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 infor=
mation is generated.


Workaround : this stops happening when jit_above_cost is set to -1 in postg=
resql.conf=C2=A0=C2=A0 After changing this setting the problem stops.

This was reported in the past for MacOS, and the same workaround was report=
ed to work in that case.

------=_Part_9759141_267692230.1724114730508
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<html><head></head><body><div class=3D"ydp6bb93bf7yahoo-style-wrap" style=
=3D"font-family:Helvetica Neue, Helvetica, Arial, sans-serif;font-size:16px=
;"><div dir=3D"ltr" data-setdir=3D"false">postgres version:&nbsp; <span>Pos=
tgreSQL 15.8 (Raspbian 15.8-0+deb12u1) on arm-unknown-linux-gnueabihf, comp=
iled by gcc (Raspbian 12.2.0-14+rpi1)</span></div><div dir=3D"ltr" data-set=
dir=3D"false"><span>OS:&nbsp; <span>Linux pi 6.6.31+rpt-rpi-v6 #1 Raspbian =
1:6.6.31-1+rpt1 (2024-05-29) armv6l GNU/Linux</span><br></span></div><div d=
ir=3D"ltr" data-setdir=3D"false"><div dir=3D"ltr" data-setdir=3D"false"><br=
></div><div dir=3D"ltr" data-setdir=3D"false">CPU: (/proc/cpuinfo)<br></div=
><div><br></div><div>name&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; : ARMv6-compatible =
processor rev 7 (v6l)<br>BogoMIPS&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;=
 : 697.95<br>Features&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; : half thum=
b fastmult vfp edsp java tls<br>CPU implementer : 0x41<br>CPU architecture:=
 7<br>CPU variant&nbsp;&nbsp;&nbsp;&nbsp; : 0x0<br>CPU part&nbsp;&nbsp;&nbs=
p;&nbsp;&nbsp;&nbsp;&nbsp; : 0xb76<br>CPU revision&nbsp;&nbsp;&nbsp; : 7<br=
><br>Hardware&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; : BCM2835<br>Revisi=
on&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; : 000e<br>Serial&nbsp;&nbsp;&n=
bsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; : 00000000d75e7c8b<br>Model&nbsp;&=
nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; : Raspberry Pi Model =
B Rev 2</div></div><div dir=3D"ltr" data-setdir=3D"false"><br></div><div di=
r=3D"ltr" data-setdir=3D"false">issue: when issuing a large query ,&nbsp; e=
.g.&nbsp;&nbsp; "Select count(*) from temperature" on a table with &gt; 150=
M rows the system crashes.</div><div dir=3D"ltr" data-setdir=3D"false"><br>=
</div><div dir=3D"ltr" data-setdir=3D"false"><span>&nbsp; example error:&nb=
sp;&nbsp; LOG:&nbsp; server process (PID 20037) was terminated by signal 4:=
 Illegal instruction</span></div><div dir=3D"ltr" data-setdir=3D"false"><sp=
an>&nbsp; no more diagnostic information is generated.<br></span></div><div=
 dir=3D"ltr" data-setdir=3D"false"><span><br></span></div><div dir=3D"ltr" =
data-setdir=3D"false"><span><br></span></div><div dir=3D"ltr" data-setdir=
=3D"false"><span>Workaround : this stops happening when jit_above_cost is s=
et to -1 in postgresql.conf&nbsp;&nbsp; After changing this setting the pro=
blem stops.<br></span></div><div dir=3D"ltr" data-setdir=3D"false"><span><b=
r></span></div><div dir=3D"ltr" data-setdir=3D"false"><span>This was report=
ed in the past for MacOS, and the same workaround was reported to work in t=
hat case.<br></span></div></div></body></html>
------=_Part_9759141_267692230.1724114730508--