From owner-freebsd-questions Mon Jan 19 22:18:32 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id WAA06729 for questions-outgoing; Mon, 19 Jan 1998 22:18:32 -0800 (PST) (envelope-from owner-freebsd-questions@FreeBSD.ORG) Received: from ns.itep.ru (ns.itep.ru [193.124.224.35]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id WAA06572 for ; Mon, 19 Jan 1998 22:13:56 -0800 (PST) (envelope-from Fedor.Gubarev@itep.ru) Received: from raven.itep.ru (raven.itep.ru [193.124.227.26]) by ns.itep.ru (8.8.5-MVC-ITEP-100697/8.8.5) with ESMTP id JAA25761 for ; Tue, 20 Jan 1998 09:07:55 +0300 (GMT+0300) Received: from localhost (gubarev@localhost) by raven.itep.ru (8.8.7/8.8.7) with SMTP id JAA01498 for ; Tue, 20 Jan 1998 09:23:05 +0300 (MSK) (envelope-from Fedor.Gubarev@itep.ru) X-Authentication-Warning: raven.itep.ru: gubarev owned process doing -bs Date: Tue, 20 Jan 1998 09:23:05 +0300 (MSK) From: Fedor Gubarev X-Sender: gubarev@raven.itep.ru To: freebsd-questions@FreeBSD.ORG Subject: mount_umap and system crash. Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-questions@FreeBSD.ORG Precedence: bulk Hi, I have a problem with mount_umap command -- every time it results in crash of the system (2.2.5 release)... I don't know what's this - it really looks like a bug. Symptoms: 1) Trying to include support of NULLFS and UMAPFS in the kernel configuration file, I get the following: Options: options NULLFS options UMAPFS options SAFETY work fine, but with "options NULLFS_DIAGNOSTIC" or "options UMAPFS_DIAGNOSTIC" compilation of the kernel fails (although config reports no problems). 2) The usage of mount_umap reported in "man 8 mount_umap" is different from the actual usage of this command. Compare: # man 8 mount_umap mount_umap [-o options] target mount-point uid-mapfile gid-mapfile #mount_umap usage: mount_umap [-o options] -u usermap -g groupmap target_fs mount_point 3) A simplest way to crash the system using mount_umap is: with uid-mapfile and gid-mapfile consisting of a single line "0" do: # mount_umap -u uid-mapfile -g gid-mapfile /var /mnt With this command everything is OK untill you try: # umount /mnt Immediate system crash follows with some kernel message. 4) Another way to crash: suppose you may mount NFS file system: #mount -t nfs SomeHost:/somefs /home/somefs Now with uid-mapfile, gid-mapfile as in 3) do: # mount_umap -u uid-mapfile -g gid-mapfile /home/somefs /mnt Immediate crash follows after the command like: #ls -l /mnt ------------------ Please, help, since I don't understand what is going wrong. For everybody who are interested in I may give complete kernel output in the case 3)-4) or any other needed details (but this means one more crash....) Thanks for your attention, Gubarev Fedor. Fedor.Gubarev@itep.ru