From owner-freebsd-stable@FreeBSD.ORG Sat Jan 14 10:59:44 2012 Return-Path: Delivered-To: freebsd-stable@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 78F99106564A for ; Sat, 14 Jan 2012 10:59:44 +0000 (UTC) (envelope-from hlh@restart.be) Received: from tignes.restart.be (tignes.restart.be [IPv6:2001:41d0:2:56bf:0:1::]) by mx1.freebsd.org (Postfix) with ESMTP id 056C28FC12 for ; Sat, 14 Jan 2012 10:59:44 +0000 (UTC) Received: from restart.be (avoriaz.tunnel.bel [IPv6:2001:41d0:2:56bf:1:ffff::]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (Client CN "smtp.restart.be", Issuer "CA master" (verified OK)) by tignes.restart.be (Postfix) with ESMTPS id 065DB15B9A; Sat, 14 Jan 2012 11:59:42 +0100 (CET) X-DKIM: OpenDKIM Filter v2.4.1 tignes.restart.be 065DB15B9A DKIM-Signature: v=1; a=rsa-sha256; c=simple/simple; d=restart.be; s=tignes; t=1326538783; bh=pJIwaQb7uJ4qJ4dCFA8N6zfJmFeHtMSD3kmZryh9s4I=; h=Message-ID:Date:From:MIME-Version:To:CC:Subject:References: In-Reply-To:Content-Type:Content-Transfer-Encoding; z=Message-ID:=20<4F11601C.7030409@restart.be>|Date:=20Sat,=2014=20J an=202012=2011:59:40=20+0100|From:=20Henri=20Hennebert=20|MIME-Version:=201.0|To:=20Jeremy=20Chadwick=20|CC:=20freebsd-stable@freebsd.org|Subject:=20Re:=20F reeBSD=209=20&=20recompile=20ports|References:=20=20=20<4F10 009A.7000307@FreeBSD.org>=20<4F100D9E.80807@FreeBSD.org>=20=20<4 F10B1AD.9000400@FreeBSD.org>=20<4F10B6C9.8010607@FreeBSD.org>=20<4 F1140DD.4040300@infracaninophile.co.uk>=20<4F1158EC.8070803@restar t.be>=20<20120114103741.GA1936@icarus.home.lan>|In-Reply-To:=20<20 120114103741.GA1936@icarus.home.lan>|Content-Type:=20text/plain=3B =20charset=3DISO-8859-1|Content-Transfer-Encoding:=207bit; b=kS1vwq8TsFZjZwWEBOf6lJUH3QdJ5ybtF1x+DJu516ZEQ+y0gb7n0gMEIXRAAU2ft KXTF1tDyY6PSnapNJqGDxoHeJbxpXmTjbWNdMCcOObUBI+YE4mih6ucGOkQcNlMGcH e+OPxBdNfeo+5uMOmg2gc7NYwn89yswa+b81nhTQ= Received: from meribel.restart.bel (meribel.restart.bel [IPv6:2001:41d0:2:56bf:1:8::]) (authenticated bits=0) by restart.be (8.14.5/8.14.5) with ESMTP id q0EAxebs038581 (version=TLSv1/SSLv3 cipher=DHE-RSA-CAMELLIA256-SHA bits=256 verify=OK); Sat, 14 Jan 2012 11:59:41 +0100 (CET) (envelope-from hlh@restart.be) Message-ID: <4F11601C.7030409@restart.be> Date: Sat, 14 Jan 2012 11:59:40 +0100 From: Henri Hennebert User-Agent: Mozilla/5.0 (X11; FreeBSD i386; rv:8.0) Gecko/20111212 Thunderbird/8.0 MIME-Version: 1.0 To: Jeremy Chadwick References: <4F10009A.7000307@FreeBSD.org> <4F100D9E.80807@FreeBSD.org> <4F10B1AD.9000400@FreeBSD.org> <4F10B6C9.8010607@FreeBSD.org> <4F1140DD.4040300@infracaninophile.co.uk> <4F1158EC.8070803@restart.be> <20120114103741.GA1936@icarus.home.lan> In-Reply-To: <20120114103741.GA1936@icarus.home.lan> X-Enigmail-Version: undefined OpenPGP: id=D351A503 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Cc: freebsd-stable@freebsd.org Subject: Re: FreeBSD 9 & recompile ports X-BeenThere: freebsd-stable@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Production branch of FreeBSD source code List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sat, 14 Jan 2012 10:59:44 -0000 On 01/14/2012 11:37, Jeremy Chadwick wrote: > On Sat, Jan 14, 2012 at 11:29:00AM +0100, Henri Hennebert wrote: >> On 01/14/2012 09:46, Matthew Seaman wrote: >>> On 13/01/2012 22:57, Andriy Gapon wrote: >>>> But if the appropriate misc/compatX port is installed, then >>>> those libraries do actually exist and the system should be fully >>>> usable... Modulo the compat libraries not working with the new >>>> kernel as Kostik has pointed out. >>> >>> As soon as you update or install an application after this point, >>> you are likely to end up with an application that tries to >>> dynamically link two different versions of the same shlib, and >>> that is a recipe for tears-before-bedtime. >> >> This /etc/libmap.conf help me greatly when I reinstall all my ports -------> ^^^^^^^^^^^^^^^^^^^^^^^^ >> after 9.0-BETA2 and make delete-old-libs: >> >> libsbuf.so.5 libsbuf.so.6 >> libz.so.5 libz.so.6 >> libutil.so.8 libutil.so.9 >> libcam.so.5 libcam.so.6 >> libpcap.so.7 libpcap.so.8 >> libufs.so.5 libufs.so.6 >> libbsnmp.so.5 libbsnmp.so.6 >> libdwarf.so.2 libdwarf.so.3 >> libopie.so.6 libopie.so.7 >> librtld_db.so.1 librtld_db.so.2 >> libtacplus.so.4 libtacplus.so.5 > > This is very, VERY, ***VERY*** dangerous. Apparently nobody has > explained why, so I will: > > When a linked library version number (N of libfoo.so.N) increases or > changes, it indicates there are API/ABI changes to the library. There > is absolutely ZERO guarantee that calling semantics are the same, that > function arguments (thus stack order) are the same, or that structures > used internally by the library are the same. The effects of this can be > devastating -- if you're lucky it'll consist of just "missing symbol", > but it can be a lot worse. The TL;DR version is: there is absolutely > ZERO guarantee that the internal operations and calling semantics of the > libraries are identical. > > Folks reading this thread, PLEASE do not follow the above advice and > leave your system running in that kind of state. Instead of being lazy, I don't want to argue too much, but you don't read me correctly. I just do this during the time I REINSTALL ALL PORTS and then I delete /etc/libmap.conf, of course, I'm not crazy! > rebuild all your ports from scratch or pull down new binary copies > (pkg_add -r ...) for the version of the OS you're running. Doug and I > have the same opinion when it comes to this situation, and it's based > purely on experience. Schedule downtime, spend an afternoon rebuilding > things, whatever -- just do it the Right Way(tm) please. Otherwise > you're creating a lot of support hassle when it comes to trying to > diagnose why some program on your system "behaves oddly" -- weeks go by, > "oh, libmap.conf..." >