From owner-freebsd-current Sun Jun 7 16:00:06 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id QAA19046 for freebsd-current-outgoing; Sun, 7 Jun 1998 16:00:06 -0700 (PDT) (envelope-from owner-freebsd-current@FreeBSD.ORG) Received: from awfulhak.org (awfulhak.force9.co.uk [195.166.136.63]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id PAA18971 for ; Sun, 7 Jun 1998 15:59:51 -0700 (PDT) (envelope-from brian@Awfulhak.org) Received: from gate.lan.awfulhak.org (localhost [127.0.0.1]) by awfulhak.org (8.8.8/8.8.8) with ESMTP id XAA27931; Sun, 7 Jun 1998 23:33:57 +0100 (BST) (envelope-from brian@gate.lan.awfulhak.org) Message-Id: <199806072233.XAA27931@awfulhak.org> X-Mailer: exmh version 2.0.1 12/23/97 To: John Polstra cc: Brian Somers , toj@gorillanet.gorilla.net, current@FreeBSD.ORG Subject: Re: IP Packet Aliasing Broke? In-reply-to: Your message of "Sun, 07 Jun 1998 14:49:22 PDT." <199806072149.OAA19657@austin.polstra.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Sun, 07 Jun 1998 23:33:56 +0100 From: Brian Somers Sender: owner-freebsd-current@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG > > The bloat I was concerned with was more to do with the installation > > disk footprint rather than the in-core footprint > > Ah, sorry, I wasn't considering that piece of the puzzle. > > Well, it seems like there are several ways you could go with it. > Count on a symlink named "libalias.so" in some well-known place. Or, > make the hard-coded path to the library overridable from the ppp > Makefile. Or, have ppp search a few likely directories to find the > pathname for dlopen. Or, muck around in src/release/Makefile. I think I'll battle src/release/Makefile - I have to anyway (the ppp binary that shipped with 2.2.6 was wrong because I got my last update to that Makefile wrong!). Once the RELEASE_CRUNCH definition works, it's trivial to add to the .if defined(NOALIAS) || defined(RELEASE_CRUNCH) bit of the Makefile and have libalias loaded normally (if at all). > For that matter, I could probably learn to live with the dlfind() > idea if it just always found the highest minor number where minor > numbers existed at all: > > const char *dlfind(const char *name, int version); > > I think the leading "lib" part of the name should be implicit, so > that the call would be like: const char *path = dlfind("alias", 2); Agreed. It might be nice to have this anyway (even if it's not required by ppp)..... does anything else use dlopen() ? In any case, it should probably be left 'till the a.out/elf changes have solidified. > John -- Brian , , Don't _EVER_ lose your sense of humour.... To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message