From owner-freebsd-emulation@FreeBSD.ORG Tue Apr 29 20:26:41 2008 Return-Path: Delivered-To: freebsd-emulation@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 59697106568E for ; Tue, 29 Apr 2008 20:26:41 +0000 (UTC) (envelope-from jhein@timing.com) Received: from Daffy.timing.com (mx1.timing.com [206.168.13.218]) by mx1.freebsd.org (Postfix) with ESMTP id 06E268FC14 for ; Tue, 29 Apr 2008 20:26:40 +0000 (UTC) (envelope-from jhein@timing.com) Received: from gromit.timing.com (gromit.timing.com [206.168.13.209]) by Daffy.timing.com (8.13.1/8.13.1) with ESMTP id m3TKQ4tP025404; Tue, 29 Apr 2008 14:26:39 -0600 (MDT) (envelope-from jhein@timing.com) Received: from gromit.timing.com (localhost [127.0.0.1]) by gromit.timing.com (8.14.2/8.14.2) with ESMTP id m3TKQ26r052917; Tue, 29 Apr 2008 14:26:02 -0600 (MDT) (envelope-from jhein@gromit.timing.com) Received: (from jhein@localhost) by gromit.timing.com (8.14.2/8.14.2/Submit) id m3TKQ2CU052914; Tue, 29 Apr 2008 14:26:02 -0600 (MDT) (envelope-from jhein) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <18455.33882.355026.28602@gromit.timing.com> Date: Tue, 29 Apr 2008 14:26:02 -0600 From: John E Hein To: Boris Samorodov In-Reply-To: <94009951@ipt.ru> References: <48164CD5.9030109@gmail.com> <790a9fff0804281537p27003437p10ff1d9bf7c7bff@mail.gmail.com> <20080429045001.GW18958@deviant.kiev.zoral.com.ua> <42172269@ipt.ru> <20080429101459.GX18958@deviant.kiev.zoral.com.ua> <95766063@serv3.int.kfs.ru> <18455.7999.318965.875282@gromit.timing.com> <94009951@ipt.ru> X-Mailer: VM 7.19 under Emacs 22.1.1 X-Virus-Scanned: ClamAV version 0.91.2, clamav-milter version 0.91.2 on Daffy.timing.com X-Virus-Status: Clean Cc: freebsd-emulation@freebsd.org Subject: Re: linux_base-f8 giving me guff X-BeenThere: freebsd-emulation@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Development of Emulators of other operating systems List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 29 Apr 2008 20:26:41 -0000 Boris Samorodov wrote at 22:13 +0400 on Apr 29, 2008: > We have packages for that purpose... Or I didn't understand you. Q: How do packages get built? A: By running 'make package' which does a 'make install' So, the problem Kostik described still exists. When does the ldconfig really need to be run? At package build time or at pkg_add time? > > Also note that /etc/rc.d/abi runs /compat/linux/sbin/ldconfig. > > ...which will run while booting? That is not enough if you just > install a linux port. I think I don't understand your point. As you know, you currently need linux.ko installed and linux ldconfig setup right in order to run certain linux ports that use linux shared libs. /etc/rc.d/abi does both of those things (on boot or when run manually). You need linux.ko installed to run linux ldconfig which some linux ports do if they install linux shared libs. Kostik is requesting a [non-default] way to tell the ports infrastructure not to die if I want to install a linux port without linux.ko loaded. There were some concerns expressed about just warning if ldconfig doesn't get run because that would leave the ldconfig cache possibly set incorrectly. My point is that even if the linux ldconfig's cache is not right on a port install, it will be corrected the next time /etc/rc.d/abi is run (to that end, it's better than manually running 'sudo kldload linux'). So let's say you are installing a port to a chroot on removable media that is intended to be installed on another box - a useful mode of operation I hope you agree. The risk of having a incorrectly populated ldconfig cache is mitigated by the /etc/rc.d/abi script that will be run when the destination box is booted. > > At least add a knob (ala DISABLE_VULNERABILITIES) for when the user > > knows better. > > For packaging needs I suppose to use (an overwritable) variable > LINUX_OSRELEASE. It will be introduced at bsd.linux-apps.mk together > with other linux-f8 infrastructure ports. See the recent emulation@ > archives for more details. Yes, I know - I've glanced at it and am glad there's work in that direction. Thanks! How does LINUX_OSRELEASE help avoid running linux ldconfig on make install?