Date: Thu, 6 Feb 1997 01:22:09 -0500 (EST) From: StevenR362@aol.com To: msmith@atrad.adelaide.edu.au, rcarter@consys.com Cc: current@freebsd.org Subject: Re: linux ELF codine no go on 2.2 Gamma Message-ID: <970206010953_818697893@emout05.mail.aol.com>
next in thread | raw e-mail | index | archive | help
In a message dated 97-02-05 20:59:53 EST, msmith@atrad.adelaide.edu.au (Michael Smith) writes: > > Russell L. Carter stands accused of saying: > > > > I've been trying to get codine 4 (http://www.genias.de/) > > for linux ELF going with the following result: > > > > $ cod_qmaster > > ELF binary type not known > > Abort trap > > $ file cod_qmaster > > cod_qmaster: ELF 32-bit LSB executable, Intel 80386, version 1 > > It's a static ELF binary; without the path of the linker to use as hint, > there's no way to tell which ABI it requires. Try > > # brandelf cod_qmaster > # brandelf -t Linux cod_qmaster > > > on the other hand, Wingz seems to be happy as a bug, and > > file reports the same: > > > > $ file /usr/X11/lib/X11/Wingz/bin/Wingz > > /usr/X11/lib/X11/Wingz/bin/Wingz: ELF 32-bit LSB executable, Intel 80386, > > version 1 > > It's linked shared, so the execution class code can guess what it is. > > > Russell I've been wondering if there is any way to abuse the filesystem to identify Linux, SCO, and BSDI binaries. Why not make the assumption in the kernel that any binary loaded from /compat/[linux,sco,bsdi] is an elf binary of the type denoted by the subdirectory. This would eliminate any need to brand the binaries. You would just install them under the appropriate subdirectory of /compat. This makes the implicit and reasonable assumption that any elf binary not under /compat is a native FreeBSD one. STeve
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?970206010953_818697893>