From owner-cvs-bin  Wed Dec 10 22:16:28 1997
Return-Path: <owner-cvs-bin>
Received: (from root@localhost)
          by hub.freebsd.org (8.8.7/8.8.7) id WAA07163
          for cvs-bin-outgoing; Wed, 10 Dec 1997 22:16:28 -0800 (PST)
          (envelope-from owner-cvs-bin)
Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19])
          by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id WAA07144;
          Wed, 10 Dec 1997 22:16:09 -0800 (PST)
          (envelope-from bde@zeta.org.au)
Received: (from bde@localhost) by godzilla.zeta.org.au (8.8.7/8.6.9) id RAA27951; Thu, 11 Dec 1997 17:08:50 +1100
Date: Thu, 11 Dec 1997 17:08:50 +1100
From: Bruce Evans <bde@zeta.org.au>
Message-Id: <199712110608.RAA27951@godzilla.zeta.org.au>
To: cvs-all@FreeBSD.ORG, cvs-bin@FreeBSD.ORG, cvs-committers@FreeBSD.ORG,
        eivind@FreeBSD.ORG
Subject: Re: cvs commit: src/bin/mv mv.c
Sender: owner-cvs-bin@FreeBSD.ORG
X-Loop: FreeBSD.org
Precedence: bulk

>  Modified files:
>    bin/mv               mv.c 
>  Log:
>  Merge from OpenBSD:
>  > Error out if someone tries to mv a mount point.  Old behavior was to
>  > move all files contained in the mounted filesystem to the dest. dir
>  > which could be quite nasty.  Personally, I think rename(2) should
>  > return EPERM or EINVAL instead of EXDEV.
>  
>  Obtained from: OpenBSD mv.c rev 1.6 by Todd Miller <millert@openbsd.org>

This seems to break POSIX conformance.  According to a POSIX.2 draft,
mv(1) shall attempt to move the entire subtree if rename(2) returns
EXDEV, and according to POSIX.1-1990, rename(2) shall return EXDEV for
cross-device moves (unless certain other error conditions not related
to cross-device moves occur).

Moving entire subtrees is probably wrong in all cases.  Subtrees may be
huge, and they may contain mount points even if the root is not a mount
point.

Bruce