Date: Thu, 22 Feb 2001 23:50:35 -0800 From: Kris Kennaway <kris@obsecurity.org> To: Alfred Perlstein <bright@wintelcom.net> Cc: Kris Kennaway <kris@obsecurity.org>, Marcel Moolenaar <marcel@cup.hp.com>, arch@FreeBSD.ORG Subject: Re: sysctl kern.fallback_elf_brand Message-ID: <20010222235035.A1656@mollari.cthul.hu> In-Reply-To: <20010222234457.D8663@fw.wintelcom.net>; from bright@wintelcom.net on Thu, Feb 22, 2001 at 11:44:57PM -0800 References: <3A960EF8.75C3FC53@cup.hp.com> <20010222233800.A1394@mollari.cthul.hu> <20010222234457.D8663@fw.wintelcom.net>
next in thread | previous in thread | raw e-mail | index | archive | help
--YZ5djTAD1cGYuMQK Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On Thu, Feb 22, 2001 at 11:44:57PM -0800, Alfred Perlstein wrote: > * Kris Kennaway <kris@obsecurity.org> [010222 23:38] wrote: > > On Thu, Feb 22, 2001 at 11:19:20PM -0800, Marcel Moolenaar wrote: > > > Hi, > > >=20 > > > I'm worried about the default value of the kern.fallback_elf_brand sy= ctl > > > (=3D 9 -> FreeBSD). It basicly tells the kernel that binaries without= any > > > branding are FreeBSD binaries. Since our binaries are always branded > > > (AFAICT), this seems to me as the wrong default. > > >=20 > > > One problem with this is that unbranded static Linux binaries are > > > executed as FreeBSD native binaries and there's a high chance of them > > > rebooting the machine if run as root. > > >=20 > > > I think we need to disable the fallback ELF branding when no ABI > > > compatibility module is loaded. Otherwise we can set the fallback to = the > > > one ABI module, or when multiple are loaded, the first. In the latter > > > case, the first may not be the preferred one, so we probably need to > > > have a bit more tuning than simply selecting the first. > > >=20 > > > Of course, we can also set the default to 3 (=3DLinux) under the > > > assumption that the Linuxulator is the most frequently used ABI modul= e. > > >=20 > > > Thoughts? > >=20 > > I've run into the unbranded Linux binary reboot before..very > > annoying. I agree the default should be changed. >=20 > Why does this happen? Does the exec code freak out if the default > isn't present, or does some common syscall just happen to map to > Linux's reboot syscall? Other way around. A common Linux syscall maps to the FreeBSD reboot syscall, so if the binary is unbranded the syscalls are interpreted using the FreeBSD table, and the user is left looking very surprised. Kris --YZ5djTAD1cGYuMQK Content-Type: application/pgp-signature Content-Disposition: inline -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.0.4 (FreeBSD) Comment: For info see http://www.gnupg.org iD8DBQE6lhZLWry0BWjoQKURApc7AKCA3+eJuOL2JaWcOjfTVXTruqCDJQCfUzdH OLsW9gV1VG2jUau5kwb9FX0= =FxRa -----END PGP SIGNATURE----- --YZ5djTAD1cGYuMQK-- To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-arch" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20010222235035.A1656>