Your message dated Wed, 17 Feb 2010 18:55:19 +0100 with message-id <20100217175519.GD17517@inutil.org> and subject line Re: kernel limitation has caused the Debian Bug report #547338, regarding rename()/link() really ought to be able to deal with bind-mounts to be marked as done. This means that you claim that the problem has been dealt with. If this is not the case it is now your responsibility to reopen the Bug report if necessary, and/or fix the problem forthwith. (NB: If you are a system administrator and have no idea what this message is talking about, this may indicate a serious mail system misconfiguration somewhere. Please contact owner@bugs.debian.org immediately.) -- 547338: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=547338 Debian Bug Tracking System Contact owner@bugs.debian.org with problems
--- Begin Message ---
- To: Debian Bug Tracking System <submit@bugs.debian.org>
- Subject: rename()/link() think bind-mounts are different devices
- From: martin f krafft <madduck@debian.org>
- Date: Fri, 18 Sep 2009 20:06:50 +0200
- Message-id: <20090918180650.GA20825@lapse.rw.madduck.net>
Package: libc6 Version: 2.9-26 Severity: normal File: /lib/libc-2.9.so This may be a kernel bug, if so, please reassign. In the following, /flics is a bind-mount of /srv/flics, as exemplified by the stat output (device number and inode identical). /bin/mv ends up with copy-delete (new inode) rather than rename: vizier:~|master|% stat -c %D/%i /flics /srv/flics fd05/480513 fd05/480513 vizier:~|master|% touch /flics/foo vizier:~|master|% stat -c %D/%i /flics/foo fd05/206081 vizier:~|master|% mv /flics/foo /srv/flics/bar vizier:~|master|% stat -c %D/%i /srv/flics/bar fd05/206082 The problem is this (thanks, strace): rename("/flics/foo", "/srv/flics/bar") = -1 EXDEV (Invalid cross-device link) so rename() does not properly detect that the source and destination directories are actually the same. The problem seems identical for link(). -- System Information: Debian Release: squeeze/sid APT prefers unstable APT policy: (500, 'unstable'), (500, 'stable'), (1, 'experimental') Architecture: i386 (i686) Kernel: Linux 2.6.30-1-686 (SMP w/1 CPU core) Locale: LANG=en_GB, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Versions of packages libc6 depends on: ii libc-bin 2.9-26 GNU C Library: Binaries ii libgcc1 1:4.4.1-4 GCC support library Versions of packages libc6 recommends: ii libc6-i686 2.9-26 GNU C Library: Shared libraries [i Versions of packages libc6 suggests: ii debconf [debconf-2.0] 1.5.27 Debian configuration management sy pn glibc-doc <none> (no description available) ii locales 2.9-26 GNU C Library: National Language ( -- debconf-show failed -- .''`. martin f. krafft <madduck@d.o> Related projects: : :' : proud Debian developer http://debiansystem.info `. `'` http://people.debian.org/~madduck http://vcs-pkg.org `- Debian - when you have better things to do than fixing systemsAttachment: digital_signature_gpg.asc
Description: Digital signature (see http://martin-krafft.net/gpg/)
--- End Message ---
--- Begin Message ---
- To: martin f krafft <madduck@debian.org>
- Cc: 547338-done@bugs.debian.org
- Subject: Re: kernel limitation
- From: Moritz Muehlenhoff <jmm@inutil.org>
- Date: Wed, 17 Feb 2010 18:55:19 +0100
- Message-id: <20100217175519.GD17517@inutil.org>
- In-reply-to: <20090918181625.GA21061@lapse.rw.madduck.net>
- References: <20090918181625.GA21061@lapse.rw.madduck.net>
martin f krafft wrote: > reassign 547338 linux-2.6 > severity 547338 wishlist > retitle 547338 rename()/link() really ought to be able to deal with bind-mounts > thanks > > This is a known limitation (see rename(2) manpage) and I am sorry > for filing it prematurely. I am reassigning to kernel as a wishlist > because it really can't (or ought not) be hard to have this work. Closing, this only clutters the BTS. If you want to see this changed, start a discussion on linux-kernel on how this can be implemented in a correct manner. Cheers, Moritz
--- End Message ---