Date: Sat, 17 Nov 2012 17:31:54 +0100 From: Ed Schouten <ed@80386.nl> To: Eitan Adler <lists@eitanadler.com> Cc: rlucia@iscanet.com, freebsd-jail@freebsd.org Subject: Re: bin/32828: [jail] w(1) incorrectly handles stale utmp slots with jail Message-ID: <CAJOYFBAFSE%2BJuPGjF9qcrQc1WieqFNdXE3ptr989O4cfGZyDNw@mail.gmail.com> In-Reply-To: <CAF6rxgnBREuufuoAZmGazzYrxOT7f4OVjE1C7KLfwNf6i%2B81CA@mail.gmail.com> References: <201211171312.qAHDChWT093266@freefall.freebsd.org> <CAF6rxgnBREuufuoAZmGazzYrxOT7f4OVjE1C7KLfwNf6i%2B81CA@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Hello Eitan, 2012/11/17 Eitan Adler <lists@eitanadler.com>: > Was the fix MFCed? Well, yes and no. There are actually three separate changes here: 1. The visibility of pseudo-terminals. On FreeBSD 8 and later, pseudo-terminals are only shown in devfs if they belong to your jail or one of its descendants. 2. If you set cleanvar_enable="YES" your utx.active file will already be discarded on startup, meaning any stale entries won't survive a reboot. 3. r231534 moves the unconditional initialisation of utx.active with a BOOT_TIME entry out of init(8) into an rc script, so you don't need cleanvar_enable="YES". I'm not planning on MFCing the latter, as FreeBSD 9's utmpx does not write BOOT_TIME entries in utx.active to begin with, so merging back r231534 is useless. Given the fact that the first two changes already effectively fix this issue, I consider this bug fixed. -- Ed Schouten <ed@80386.nl>
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAJOYFBAFSE%2BJuPGjF9qcrQc1WieqFNdXE3ptr989O4cfGZyDNw>