From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from outpost1.zedat.fu-berlin.de (outpost1.zedat.fu-berlin.de [130.133.4.66]) by sourceware.org (Postfix) with ESMTPS id 2B8BB3858D32 for ; Mon, 2 Jan 2023 11:44:44 +0000 (GMT) Received: from inpost2.zedat.fu-berlin.de ([130.133.4.69]) by outpost.zedat.fu-berlin.de (Exim 4.95) with esmtps (TLS1.3) tls TLS_AES_256_GCM_SHA384 (envelope-from ) id 1pCJFA-002eu1-Le; Mon, 02 Jan 2023 12:44:40 +0100 Received: from p5b13ac85.dip0.t-ipconnect.de ([91.19.172.133] helo=[192.168.178.81]) by inpost2.zedat.fu-berlin.de (Exim 4.95) with esmtpsa (TLS1.3) tls TLS_AES_128_GCM_SHA256 (envelope-from ) id 1pCJFA-000ula-FO; Mon, 02 Jan 2023 12:44:40 +0100 Message-ID: <999879fc-9cb3-1efa-165b-36a5418f2e41@physik.fu-berlin.de> Date: Mon, 2 Jan 2023 12:44:40 +0100 MIME-Version: 1.0 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101 Thunderbird/102.6.1 From: John Paul Adrian Glaubitz Subject: Update on the glibc segfault issue on Alpha To: "debian-alpha@lists.debian.org" Content-Language: en-US Cc: gentoo-alpha@lists.gentoo.org, GNU C Library Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit X-Original-Sender: glaubitz@physik.fu-berlin.de X-Originating-IP: 91.19.172.133 X-Spam-Status: No, score=1.3 required=5.0 tests=BAYES_00,JMQ_SPF_NEUTRAL,KAM_DMARC_STATUS,RCVD_IN_BARRACUDACENTRAL,RCVD_IN_MSPIKE_H3,RCVD_IN_MSPIKE_WL,SPF_HELO_NONE,SPF_PASS,TXREP,URIBL_BLACK autolearn=no autolearn_force=no version=3.4.6 X-Spam-Level: * X-Spam-Checker-Version: SpamAssassin 3.4.6 (2021-04-09) on server2.sourceware.org List-Id: Hello! Just a quick update on the glibc issue we're seeing on Alpha! My previous bisecting result was wrong and the segfault I ran into was the result of combining ld.so and libc.so files from different builds by not including all necessary paths in my LD_LIBRARY_PATH. So, my previously reported bug report was invalid [1]. However, as can be seen from the build logs, there are still clearly issues with the testsuite [2]. Further analysis showed that the problem are static binaries which segfault. Anything dynamically linked, works. This becomes obvious when building the glibc Debian package manually with the testsuite disabled and then installing these packages inside a test chroot. While the libc-bin package is being configured, ldconfig is run - a statically linked binary - and crashes with a segfault. Adhemerval from glibc upstream is aware of the problem but he has not found yet a solution to this issue as it needs to be debugged further. I will try to bisect which particular introduced the regression and file a new upstream bug report. During our discussion, Adhemerval pointed out that this change [3] might be the culprit but I have not been able to verify this yet. Adrian > [1] https://sourceware.org/bugzilla/show_bug.cgi?id=29899 > [2] https://buildd.debian.org/status/fetch.php?pkg=glibc&arch=alpha&ver=2.36-6&stamp=1669706903&raw=0 > [3] https://sourceware.org/pipermail/glibc-cvs/2020q2/069444.html -- .''`. John Paul Adrian Glaubitz : :' : Debian Developer `. `' Physicist `- GPG: 62FF 8A75 84E0 2956 9546 0006 7426 3B37 F5B5 F913