Date: Thu, 5 Sep 2002 16:46:28 +0200 From: Roman Neuhauser <neuhauser@bellavista.cz> To: Lowell Gilbert <freebsd-questions-local@be-well.no-ip.com> Cc: freebsd-questions@freebsd.org Subject: Re: mount(8) mount_union(8) and the slippery wet floor... Message-ID: <20020905144627.GL10717@freepuppy.bellavista.cz> In-Reply-To: <44bs7c1phq.fsf@be-well.ilk.org> References: <5.1.0.14.0.20020904132723.00bc28e8@mail.lusidor.nu> <20020904124832.GA15994@submonkey.net> <44y9ahecn2.fsf@be-well.ilk.org> <20020905091754.GE10717@freepuppy.bellavista.cz> <44bs7c1phq.fsf@be-well.ilk.org>
next in thread | previous in thread | raw e-mail | index | archive | help
# freebsd-questions-local@be-well.no-ip.com / 2002-09-05 10:26:57 -0400:
> Roman Neuhauser <neuhauser@bellavista.cz> writes:
> > # freebsd-questions-local@be-well.no-ip.com / 2002-09-04 16:13:53 -0400:
> > > Ceri Davies <setantae@submonkey.net> writes:
> > >
> > > > On Wed, Sep 04, 2002 at 01:46:42PM +0200, Jimmy Lantz wrote:
> > > > > Hi,
> > > > >
> > > > > My question is:
> > > > > Is the mount union mentioned in MOUNT(8) also affected by the
> > > > > slippery wet floor in MOUNT_UNION(8)??
> > > > > Or can I safely use the mount with the option union?
> > > >
> > > > They are the same thing.
> > >
> > > For the record, they are *not* the same thing.
> >
> > could you elaborate?
>
> I'm not 100% sure that I completely understand the technical details,
> but I'll give it a shot. You could go check the archives of the
> freebsd-fs list for more authoritative information on the subject.
>
> Very briefly, they implement similar functionality through different
> abstractions. mount_union is a filesystem in its own right -- it
> "stacks" on top of other filesystem types and uses their capabilities
> to perform the I/O operations, but to the system I/O code, it looks
> like a filesystem. The union mount option is visible at a higher
> level; the filesystem code itself doesn't know about the shadowing.
>
> Or something like that. At any rate, they don't seem to share any
> code.
this is interesting. I always thought that mount(8) works by
invoking mount_whatever(8) where whatever is determined by the value
passed in the -t option to mount(8). looks like this isn't the case,
then... hm:
If the type is not one of the internally known types, mount will
attempt to execute a program in /sbin/mount_XXX where XXX is
replaced by the type name. For example, nfs filesystems are
mounted by the program /sbin/mount_nfs.
And elsewhere:
For example, the mount command:
mount -t mfs -o nosuid,-N,-s=4000 /dev/dk0b /tmp
causes mount to execute the equivalent of:
/sbin/mount_mfs -o nosuid -N -s 4000 /dev/dk0b /tmp
So, this would look as though mount -t union / mount_union is a hack
defiating from the normal way these commands operate?
--
begin 666 nonexistent.vbs
FreeBSD 4.6-STABLE
4:33PM up 15 days, 22:26, 8 users, load averages: 0.10, 0.04, 0.01
end
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-questions" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20020905144627.GL10717>