Received: from malur.postgresql.org ([217.196.149.56]) by arkaria.postgresql.org with esmtps (TLS1.3:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.92) (envelope-from ) id 1nQDJS-0004MM-6e for pgsql-general@arkaria.postgresql.org; Fri, 04 Mar 2022 19:10:02 +0000 Received: from localhost ([127.0.0.1] helo=malur.postgresql.org) by malur.postgresql.org with esmtp (Exim 4.92) (envelope-from ) id 1nQDJR-0001s4-5Q for pgsql-general@arkaria.postgresql.org; Fri, 04 Mar 2022 19:10:01 +0000 Received: from magus.postgresql.org ([2a02:c0:301:0:ffff::29]) by malur.postgresql.org with esmtps (TLS1.3:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.92) (envelope-from ) id 1nQDJQ-0001pS-QB for pgsql-general@lists.postgresql.org; Fri, 04 Mar 2022 19:10:00 +0000 Received: from sss.pgh.pa.us ([66.207.139.130]) by magus.postgresql.org with esmtps (TLS1.3:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.92) (envelope-from ) id 1nQDJO-0003T6-Lb for pgsql-general@lists.postgresql.org; Fri, 04 Mar 2022 19:10:00 +0000 Received: from sss1.sss.pgh.pa.us (localhost [127.0.0.1]) by sss.pgh.pa.us (8.15.2/8.15.2) with ESMTP id 224J9qRp3969791; Fri, 4 Mar 2022 14:09:52 -0500 From: Tom Lane To: depesz@depesz.com cc: PostgreSQL General , Andres Freund Subject: Re: Interesting fail when migrating Pg from Ubuntu Bionic to Focal In-reply-to: <20220303164401.GA13056@depesz.com> References: <20220303150428.GA26036@depesz.com> <20220303151156.GB26036@depesz.com> <20220303163921.GA10776@depesz.com> <20220303164401.GA13056@depesz.com> Comments: In-reply-to hubert depesz lubaczewski message dated "Thu, 03 Mar 2022 17:44:01 +0100" MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-ID: <3969789.1646420992.1@sss.pgh.pa.us> Content-Transfer-Encoding: quoted-printable Date: Fri, 04 Mar 2022 14:09:52 -0500 Message-ID: <3969790.1646420992@sss.pgh.pa.us> List-Id: List-Help: List-Subscribe: List-Post: List-Owner: List-Archive: Archived-At: Precedence: bulk hubert depesz lubaczewski writes: > OK. Traced it back to JIT. With JIT enabled: Hah, that's useful info. Seems like it must be incorrect code generated by JIT. > versions of things that I think are relevant: > =3D$ dpkg -l | grep -E 'llvm|clang|gcc|glibc' > ii gcc 4:9.3.0-1ubuntu2 = arm64 GNU C compiler > ii gcc-10-base:arm64 10.3.0-1ubuntu1~20.04 = arm64 GCC, the GNU Compiler Collection (base package) > ii gcc-9 9.3.0-17ubuntu1~20.04 = arm64 GNU C compiler > ii gcc-9-base:arm64 9.3.0-17ubuntu1~20.04 = arm64 GCC, the GNU Compiler Collection (base package) > ii libgcc-9-dev:arm64 9.3.0-17ubuntu1~20.04 = arm64 GCC support library (development files) > ii libgcc-s1:arm64 10.3.0-1ubuntu1~20.04 = arm64 GCC support library > ii libllvm9:arm64 1:9.0.1-12 = arm64 Modular compiler and toolchain technologies, runtime lib= rary arm64, eh? I wonder if that's buggier than the Intel code paths. I tried and failed to reproduce this on Fedora 35 on aarch64, but that has what I think is a newer LLVM version: clang-13.0.0-3.fc35.aarch64 clang-libs-13.0.0-3.fc35.aarch64 clang-resource-filesystem-13.0.0-3.fc35.aarch64 gcc-11.2.1-9.fc35.aarch64 gcc-c++-11.2.1-9.fc35.aarch64 llvm-13.0.0-4.fc35.aarch64 llvm-devel-13.0.0-4.fc35.aarch64 llvm-libs-13.0.0-4.fc35.aarch64 llvm-static-13.0.0-4.fc35.aarch64 llvm-test-13.0.0-4.fc35.aarch64 Don't think I can readily install anything as old as LLVM 9 ... regards, tom lane