From: Tom Lane <tgl@sss.pgh.pa.us>
To: Paul Guyot <pguyot@kallisys.net>
cc: pgsql-bugs@lists.postgresql.org
Subject: Re: BUG #16990: Random PANIC in qemu user context
In-reply-to: <86C24765-95F7-464F-9677-B09A396A5F69@kallisys.net>
References: <16990-10b586bc699fd234@postgresql.org>
 <3714052.1619980429@sss.pgh.pa.us>
 <86C24765-95F7-464F-9677-B09A396A5F69@kallisys.net>
Comments: In-reply-to Paul Guyot <pguyot@kallisys.net>
	message dated "Sun, 02 May 2021 22:20:39 +0200"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <3723541.1619993998.1@sss.pgh.pa.us>
Content-Transfer-Encoding: quoted-printable
Date: Sun, 02 May 2021 18:19:58 -0400
Message-ID: <3723542.1619993998@sss.pgh.pa.us>
Archived-At: 
 <https://www.postgresql.org/message-id/3723542.1619993998%40sss.pgh.pa.us>
Precedence: bulk

Paul Guyot <pguyot@kallisys.net> writes:
> QEMU user emulation documentation mentions something specific to threadi=
ng on ARM.
> https://qemu.readthedocs.io/en/latest/user/main.html
>> Threading:
>> On Linux, QEMU can emulate the clone syscall and create a real host thr=
ead (with a separate virtual CPU) for each emulated thread. Note that not =
all targets currently emulate atomic operations correctly. x86 and Arm use=
 a global lock in order to preserve their semantics.

Oooh, that's pretty suggestive.  I wonder just when that "global lock" is
taken or released.  Maybe the timing of it is such that it interferes
with ever seeing our lock as free?

			regards, tom lane