Debian Bug report logs - #391319
schroot: leftover processes cause umount to fail

Toggle useless messages

Message #5 received at submit@bugs.debian.org (full text, mbox, reply):

From: "Kees Cook" <kees@outflux.net>

To: Debian Bug Tracking System <submit@bugs.debian.org>

Subject: schroot: leftover processes cause umount to fail

Date: Thu, 05 Oct 2006 17:00:48 -0700

[Message part 1 (text/plain, inline)]

Package: schroot
Version: 1.0.2-1
Severity: normal
Tags: patch


While using schroot on LVM snapshots with sbuild, I have sometimes run into
situations where build deps start up services (like cron, dbus, etc).
Once the build is finished, schroot attempts to umount and destroy the
LVM snapshot, but since there are still processes running in the chroot,
the umount fails.

Since the chroot is "over", it seems like it makes sense to kill all the
processes left in the chroot, and then reattempt to umount.  This patch
implements that.  Does this seem like a sensible change?

-- System Information:
Debian Release: testing/unstable
  APT prefers unstable
  APT policy: (500, 'unstable'), (500, 'testing')
Architecture: i386 (x86_64)
Shell:  /bin/sh linked to /bin/bash
Kernel: Linux 2.6.17-10-generic
Locale: LANG=C, LC_CTYPE=C (charmap=ANSI_X3.4-1968)

[schroot-kill-procs.patch (text/plain, attachment)]

Message #10 received at 391319@bugs.debian.org (full text, mbox, reply):

From: Kees Cook <kees@outflux.net>

To: 391319@bugs.debian.org

Subject: Re: Bug#391319: Acknowledgement (schroot: leftover processes cause umount to fail)

Date: Thu, 5 Oct 2006 18:38:35 -0700

[Message part 1 (text/plain, inline)]

Whoops, forgot the quotes around the mount location.  New patch 
attached.

-- 
Kees Cook                                            @outflux.net

[schroot-kill-procs.patch (text/plain, attachment)]

Message #15 received at 391319@bugs.debian.org (full text, mbox, reply):

From: Roger Leigh <rleigh@whinlatter.ukfsn.org>

To: Kees Cook <kees@outflux.net>

Cc: 391319@bugs.debian.org

Subject: Re: [Buildd-tools-devel] Bug#391319: schroot: leftover processes cause umount to fail

Date: Sat, 07 Oct 2006 10:54:10 +0100

[Message part 1 (text/plain, inline)]

"Kees Cook" <kees@outflux.net> writes:

> While using schroot on LVM snapshots with sbuild, I have sometimes run into
> situations where build deps start up services (like cron, dbus, etc).
> Once the build is finished, schroot attempts to umount and destroy the
> LVM snapshot, but since there are still processes running in the chroot,
> the umount fails.
>
> Since the chroot is "over", it seems like it makes sense to kill all the
> processes left in the chroot, and then reattempt to umount.  This patch
> implements that.  Does this seem like a sensible change?

Yes, it's a good idea.

One concern I have is what will happen to bind mounted filesystems, or
filesystems mounted multiple times?  If there are open files, we don't
want to kill anything /outside/ the chroot, which may also be using
the filesystem.

As an example, if I have a /srv mounted inside and outside the chroot,
and a daemon outside is using it, and a daemon inside is using it, the
umount will fail, but only the inside daemon should be killed.

Another example is a bind mounted /home.  If this fails to umount, we
might blow away all the user's processes on the entire system.

If this can be done safely, I'll be happy to apply the patch.


Regards,
Roger

-- 
  .''`.  Roger Leigh
 : :' :  Debian GNU/Linux             http://people.debian.org/~rleigh/
 `. `'   Printing on GNU/Linux?       http://gutenprint.sourceforge.net/
   `-    GPG Public Key: 0x25BFB848   Please GPG sign your mail.

[Message part 2 (application/pgp-signature, inline)]

Message #20 received at 391319@bugs.debian.org (full text, mbox, reply):

From: Kees Cook <kees@outflux.net>

To: Roger Leigh <rleigh@whinlatter.ukfsn.org>

Cc: 391319@bugs.debian.org

Subject: Re: [Buildd-tools-devel] Bug#391319: schroot: leftover processes cause umount to fail

Date: Sat, 7 Oct 2006 10:31:44 -0700

[Message part 1 (text/plain, inline)]

On Sat, Oct 07, 2006 at 10:54:10AM +0100, Roger Leigh wrote:
> One concern I have is what will happen to bind mounted filesystems, or
> filesystems mounted multiple times?  If there are open files, we don't
> want to kill anything /outside/ the chroot, which may also be using
> the filesystem.

Ah, yes.  fuser isn't as smart as I was hoping.  Here is a better patch, 
which looks for processes that were run from the chroot base dir, which 
will protect processes on mount points built with "bind".  This will 
also not kill processes that are using the chroot area but were run from 
outside the chroot.  (Causing the umounts to correctly fail.)

How does this look?

-- 
Kees Cook                                            @outflux.net

[schroot-kill-procs.patch (text/plain, attachment)]

Message #25 received at 391319@bugs.debian.org (full text, mbox, reply):

From: Kees Cook <kees@outflux.net>

To: 391319@bugs.debian.org

Subject: more improvements to kill-proc patch

Date: Tue, 24 Oct 2006 09:32:07 -0700

[Message part 1 (text/plain, inline)]

I ran into a few cases of sticky processes, so this patch is an update 
that tries to kill processes a little harder if they don't go away on 
their own.

-- 
Kees Cook                                            @outflux.net

[schroot-kill-procs.patch (text/plain, attachment)]

Message #30 received at 391319@bugs.debian.org (full text, mbox, reply):

From: Roger Leigh <rleigh@whinlatter.ukfsn.org>

To: Kees Cook <kees@outflux.net>

Cc: 391319@bugs.debian.org

Subject: Re: [Buildd-tools-devel] Bug#391319: more improvements to kill-proc patch

Date: Fri, 27 Oct 2006 23:31:26 +0100

[Message part 1 (text/plain, inline)]

Kees Cook <kees@outflux.net> writes:

> I ran into a few cases of sticky processes, so this patch is an update 
> that tries to kill processes a little harder if they don't go away on 
> their own.

Many thanks for the patch.  I hope I'll get a chance to give it some
testing this weekend.

Because schroot 1.0 is frozen for etch, I'd prefer to put this on the
trunk only at the moment.  Once it's had some more testing, I might
then consider merging it onto the schroot-1.0 branch if it has proven
itself.


Thanks again,
Roger

-- 
  .''`.  Roger Leigh
 : :' :  Debian GNU/Linux             http://people.debian.org/~rleigh/
 `. `'   Printing on GNU/Linux?       http://gutenprint.sourceforge.net/
   `-    GPG Public Key: 0x25BFB848   Please GPG sign your mail.

[Message part 2 (application/pgp-signature, inline)]

Message #35 received at 391319@bugs.debian.org (full text, mbox, reply):

From: Kees Cook <kees@outflux.net>

To: Roger Leigh <rleigh@whinlatter.ukfsn.org>

Cc: 391319@bugs.debian.org

Subject: Re: [Buildd-tools-devel] Bug#391319: more improvements to kill-proc patch

Date: Fri, 27 Oct 2006 15:46:02 -0700

On Fri, Oct 27, 2006 at 11:31:26PM +0100, Roger Leigh wrote:
> Many thanks for the patch.  I hope I'll get a chance to give it some
> testing this weekend.
> 
> Because schroot 1.0 is frozen for etch, I'd prefer to put this on the
> trunk only at the moment.  Once it's had some more testing, I might
> then consider merging it onto the schroot-1.0 branch if it has proven
> itself.

That sounds great, thank you!  I'm in no rush, I would just like to see 
it in mainline some day.  :)

-- 
Kees Cook                                            @outflux.net

Message #40 received at 391319@bugs.debian.org (full text, mbox, reply):

From: Kees Cook <kees@outflux.net>

To: 391319@bugs.debian.org

Subject: updated patch

Date: Tue, 31 Oct 2006 12:12:56 -0800

[Message part 1 (text/plain, inline)]

Oops.  Prior patch was not /bin/sh safe.  New patch attached...

-- 
Kees Cook                                            @outflux.net

[schroot-kill-procs.patch (text/plain, attachment)]

Message #45 received at 391319@bugs.debian.org (full text, mbox, reply):

From: Kees Cook <kees@outflux.net>

To: 391319@bugs.debian.org

Subject: ping

Date: Tue, 22 May 2007 21:03:06 -0700

Hello!  Any progress on incorporating this patch?  I find it very handy 
to have this patch installed so I can forcefully shutdown an schroot.

Thanks!

-- 
Kees Cook                                            @outflux.net

Message #50 received at 391319@bugs.debian.org (full text, mbox, reply):

From: Roger Leigh <rleigh@whinlatter.ukfsn.org>

To: Kees Cook <kees@outflux.net>

Cc: 391319@bugs.debian.org

Subject: Re: [Buildd-tools-devel] Bug#391319: ping

Date: Wed, 23 May 2007 20:38:15 +0100

[Message part 1 (text/plain, inline)]

Kees Cook <kees@outflux.net> writes:

> Hello!  Any progress on incorporating this patch?  I find it very
> handy to have this patch installed so I can forcefully shutdown an
> schroot.

I hope to get it in the next release.  I've been using it for a few
months, and now that etch is released it can go in.

The one problem I would like to get fixed first, however, is the delay
while killing processes, if there are no processes to kill.  This
makes the typical case of running schroot or ending a session wait for
a good few seconds.  I would like it to be ~instantaneous in the case
no processes need killing.

I think the while loop needs some optimisation, such as checking if
the process has terminated so that it can break out of the loop early.
I think simply repeating the readlink inside the loop would be
sufficient.  So instead of "while kill", so "while readlink", and then
kill inside the loop.

Another nice addition would be to print out which processes are being
killed if in verbose mode.


Regards,
Roger

-- 
  .''`.  Roger Leigh
 : :' :  Debian GNU/Linux             http://people.debian.org/~rleigh/
 `. `'   Printing on GNU/Linux?       http://gutenprint.sourceforge.net/
   `-    GPG Public Key: 0x25BFB848   Please GPG sign your mail.

[Message part 2 (application/pgp-signature, inline)]

Message #55 received at 391319@bugs.debian.org (full text, mbox, reply):

From: Kees Cook <kees@outflux.net>

To: Roger Leigh <rleigh@whinlatter.ukfsn.org>

Cc: 391319@bugs.debian.org

Subject: Re: [Buildd-tools-devel] Bug#391319: ping

Date: Wed, 23 May 2007 13:24:59 -0700

[Message part 1 (text/plain, inline)]

On Wed, May 23, 2007 at 08:38:15PM +0100, Roger Leigh wrote:
> I hope to get it in the next release.  I've been using it for a few
> months, and now that etch is released it can go in.

Excellent!  :)

> The one problem I would like to get fixed first, however, is the delay
> while killing processes, if there are no processes to kill.  This
> makes the typical case of running schroot or ending a session wait for
> a good few seconds.  I would like it to be ~instantaneous in the case
> no processes need killing.

This is already true.  If no proc items are found to kill, the function
immediately exits.

> I think the while loop needs some optimisation, such as checking if
> the process has terminated so that it can break out of the loop early.
> I think simply repeating the readlink inside the loop would be
> sufficient.  So instead of "while kill", so "while readlink", and then
> kill inside the loop.

I was trading off between a 100% CPU spin on readlink vs 1 second
granularity for each process that fails to immediately die.  As it is
written, if the process dies immediately, the "kill -0" will fail and no
waiting will be done.

> Another nice addition would be to print out which processes are being
> killed if in verbose mode.

Is this valuable?  By the time the user sees the warning, the process is
already dead.  Perhaps both the pid and the exe name?

Attached is an improved patch.

-- 
Kees Cook                                            @outflux.net

[better.diff (text/x-diff, attachment)]

Message #60 received at 391319@bugs.debian.org (full text, mbox, reply):

From: Roger Leigh <rleigh@whinlatter.ukfsn.org>

To: Kees Cook <kees@outflux.net>

Cc: 391319@bugs.debian.org

Subject: Re: [Buildd-tools-devel] Bug#391319: schroot: leftover processes cause umount to fail

Date: Mon, 28 May 2007 13:17:30 +0100

[Message part 1 (text/plain, inline)]

tags 391319 + fixed-upstream pending
thanks

Kees Cook <kees@outflux.net> writes:

> On Wed, May 23, 2007 at 08:38:15PM +0100, Roger Leigh wrote:
> This is already true.  If no proc items are found to kill, the function
> immediately exits.

Great.

>> I think the while loop needs some optimisation, such as checking if
>> the process has terminated so that it can break out of the loop early.
>> I think simply repeating the readlink inside the loop would be
>> sufficient.  So instead of "while kill", so "while readlink", and then
>> kill inside the loop.
>
> I was trading off between a 100% CPU spin on readlink vs 1 second
> granularity for each process that fails to immediately die.  As it is
> written, if the process dies immediately, the "kill -0" will fail and no
> waiting will be done.
>
>> Another nice addition would be to print out which processes are being
>> killed if in verbose mode.
>
> Is this valuable?  By the time the user sees the warning, the process is
> already dead.  Perhaps both the pid and the exe name?
>
> Attached is an improved patch.

This is fantastic.  I have separated it from 10mount, as 15killprocs
(it will always get run before 10mount on shutdown).  This is because
you might want to run it even if there are no filesystems to unmount.

I have tweaked it slightly in a few places, but it's otherwise the
same as your patch.  This is mainly using /proc/pid/root as well as
/proc/pid/exe (saves a grep invocation), and adding an additional log
message.  I also used signal names rather than numbers with kill (for
portability).  I also took the liberty of adding you to the AUTHORS
file for this contribution--I hope that's acceptable to you?

I have attached the patch I have just committed to SVN.  Is this OK,
or would you like any further changes making?


Thanks,
Roger

-- 
  .''`.  Roger Leigh
 : :' :  Debian GNU/Linux             http://people.debian.org/~rleigh/
 `. `'   Printing on GNU/Linux?       http://gutenprint.sourceforge.net/
   `-    GPG Public Key: 0x25BFB848   Please GPG sign your mail.

[schroot-15killprocs.patch (text/x-diff, inline)]

Index: debian/changelog
===================================================================
--- debian/changelog	(revision 1192)
+++ debian/changelog	(working copy)
@@ -2,8 +2,11 @@
 
   * New upstream development release.
   * debian/control: Change section of libsbuild-doc to "doc".
+  * Processes running in the chroot on stopping a session are now killed
+    by the 15killprocs setup script (Closes: #391319).  Many thanks to
+    Kees Cook for implementing this.
 
- -- Roger Leigh <rleigh@debian.org>  Sun, 20 May 2007 20:04:37 +0100
+ -- Roger Leigh <rleigh@debian.org>  Mon, 28 May 2007 13:06:05 +0100
 
 schroot (1.1.3-1) unstable; urgency=low
 
Index: AUTHORS
===================================================================
--- AUTHORS	(revision 1192)
+++ AUTHORS	(working copy)
@@ -4,6 +4,9 @@
 Andreas Bombe		<aeb@debian.org>
 	Documentation improvements.
 
+Kees Cook		<kees@outflux.net>
+	15killprocs script to kill processes.
+
 Federico Di Gregorio	<fog@debian.org>
 	Init script improvements.
 
Index: ChangeLog
===================================================================
--- ChangeLog	(revision 1192)
+++ ChangeLog	(working copy)
@@ -1,3 +1,16 @@
+2007-05-28  Roger Leigh  <rleigh@debian.org>
+
+	* debian/changelog: Close #391319.
+
+	* AUTHORS: Add Kees Cook.
+
+	* bin/schroot/setup/15killprocs: New file.  Kill processes in the
+	chroot before unmounting any filesystems.  Many thanks to Kees
+	Cook for implementing this.
+
+	* bin/schroot/setup/Makefile.am
+	(setup_SCRIPTS): Add 15killprocs
+
 2007-05-22  Roger Leigh  <rleigh@debian.org>
 
 	* bin/schroot/setup/10mount
Index: bin/schroot/setup/15killprocs
===================================================================
--- bin/schroot/setup/15killprocs	(revision 0)
+++ bin/schroot/setup/15killprocs	(revision 0)
@@ -0,0 +1,64 @@
+#!/bin/sh
+# Copyright © 2007  Kees Cook <kees@outflux.net>
+# Copyright © 2007  Roger Leigh <rleigh@debian.org>
+#
+# schroot is free software; you can redistribute it and/or modify it
+# under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 2 of the License, or
+# (at your option) any later version.
+#
+# schroot is distributed in the hope that it will be useful, but
+# WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+# General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write to the Free Software
+# Foundation, Inc., 59 Temple Place, Suite 330, Boston,
+# MA  02111-1307  USA
+
+set -e
+
+# Kill all processes that were run from within the chroot environment
+# $1: mount base location
+do_kill_all()
+{
+    if [ "$AUTH_VERBOSITY" = "verbose" ]; then
+        echo "Killing processes run inside $1"
+    fi
+    ls /proc | egrep '^[[:digit:]]+$' |
+    while read pid; do
+        exe=$(readlink /proc/"$pid"/exe || true)
+        root=$(readlink /proc/"$pid"/root || true)
+        if [ "$root" = "$1" ]; then
+            if [ "$AUTH_VERBOSITY" = "verbose" ]; then
+                echo "Killing left-over pid $pid (${exe##$1})"
+                echo "  Sending SIGTERM to pid $pid"
+            fi
+            kill -TERM "$pid" 2>/dev/null
+
+            count=0
+            max=5
+            while [ -d /proc/"$pid" ]; do
+                count=$(( $count + 1 ))
+                if [ "$AUTH_VERBOSITY" = "verbose" ]; then
+                    echo "  Waiting for pid $pid to shut down... ($count/$max)"
+                fi
+                sleep 1
+                # Wait for $max seconds for process to die before -9'ing it
+                if [ "$count" -eq "$max" ]; then
+                    if [ "$AUTH_VERBOSITY" = "verbose" ]; then
+                        echo "  Sending SIGKILL to pid $pid"
+                    fi
+                    kill -KILL "$pid" 2>/dev/null
+                    sleep 1
+                    break
+                fi
+            done
+        fi
+    done
+}
+
+if [ $1 = "setup-recover" ] || [ $1 = "setup-stop" ]; then
+    do_kill_all "$CHROOT_MOUNT_LOCATION"
+fi
\ No newline at end of file

Property changes on: bin/schroot/setup/15killprocs
___________________________________________________________________
Name: svn:executable
   + *

Index: bin/schroot/setup/Makefile.am
===================================================================
--- bin/schroot/setup/Makefile.am	(revision 1192)
+++ bin/schroot/setup/Makefile.am	(working copy)
@@ -29,6 +29,7 @@
 	05lvm		\
 	05file		\
 	10mount		\
+	15killprocs	\
 	20network	\
 	30passwd	\
 	50chrootname	\

[Message part 3 (application/pgp-signature, inline)]

Message #67 received at 391319@bugs.debian.org (full text, mbox, reply):

From: Kees Cook <kees@outflux.net>

To: Roger Leigh <rleigh@whinlatter.ukfsn.org>

Cc: 391319@bugs.debian.org

Subject: Re: [Buildd-tools-devel] Bug#391319: schroot: leftover processes cause umount to fail

Date: Mon, 28 May 2007 06:00:05 -0700

Hi Roger,

On Mon, May 28, 2007 at 01:17:30PM +0100, Roger Leigh wrote:
> This is fantastic.  I have separated it from 10mount, as 15killprocs
> (it will always get run before 10mount on shutdown).  This is because
> you might want to run it even if there are no filesystems to unmount.

Ah, yes.  Great idea.

> I have tweaked it slightly in a few places, but it's otherwise the
> same as your patch.  This is mainly using /proc/pid/root as well as
> /proc/pid/exe (saves a grep invocation), and adding an additional log
> message.  I also used signal names rather than numbers with kill (for
> portability).  I also took the liberty of adding you to the AUTHORS
> file for this contribution--I hope that's acceptable to you?

I'm honored!  Thanks, that's terrific.  I like your changes.

> I have attached the patch I have just committed to SVN.  Is this OK,
> or would you like any further changes making?

The only thing I can see to change is just a very minor performance tweak,
which is to move the exe readlink into the verbose if, just to save one
exec unless we're in verbose:

#... was here ...
+        root=$(readlink /proc/"$pid"/root || true)
+        if [ "$root" = "$1" ]; then
+            if [ "$AUTH_VERBOSITY" = "verbose" ]; then
+                exe=$(readlink /proc/"$pid"/exe || true)         #<- now here
+                echo "Killing left-over pid $pid (${exe##$1})"

Thanks!

-- 
Kees Cook                                            @outflux.net

Message #72 received at 391319@bugs.debian.org (full text, mbox, reply):

From: Roger Leigh <rleigh@whinlatter.ukfsn.org>

To: Kees Cook <kees@outflux.net>

Cc: 391319@bugs.debian.org

Subject: Re: [Buildd-tools-devel] Bug#391319: Bug#391319: schroot: leftover processes cause umount to fail

Date: Mon, 28 May 2007 14:53:25 +0100

[Message part 1 (text/plain, inline)]

Kees Cook <kees@outflux.net> writes:

> The only thing I can see to change is just a very minor performance tweak,
> which is to move the exe readlink into the verbose if, just to save one
> exec unless we're in verbose:
>
> #... was here ...
> +        root=$(readlink /proc/"$pid"/root || true)
> +        if [ "$root" = "$1" ]; then
> +            if [ "$AUTH_VERBOSITY" = "verbose" ]; then
> +                exe=$(readlink /proc/"$pid"/exe || true)         #<- now here
> +                echo "Killing left-over pid $pid (${exe##$1})"

Done!

Many thanks for all your work here.  This will be really useful, and a
lot of people have wanted something like this.


Thanks,
Roger

-- 
  .''`.  Roger Leigh
 : :' :  Debian GNU/Linux             http://people.debian.org/~rleigh/
 `. `'   Printing on GNU/Linux?       http://gutenprint.sourceforge.net/
   `-    GPG Public Key: 0x25BFB848   Please GPG sign your mail.

[Message part 2 (application/pgp-signature, inline)]

Message #77 received at 391319-close@bugs.debian.org (full text, mbox, reply):

From: Roger Leigh <rleigh@debian.org>

To: 391319-close@bugs.debian.org

Subject: Bug#391319: fixed in schroot 1.1.4-1

Date: Mon, 28 May 2007 23:02:05 +0000

Source: schroot
Source-Version: 1.1.4-1

We believe that the bug you reported is fixed in the latest version of
schroot, which is due to be installed in the Debian FTP archive:

dchroot-dsa_1.1.4-1_powerpc.deb
  to pool/main/s/schroot/dchroot-dsa_1.1.4-1_powerpc.deb
dchroot_1.1.4-1_powerpc.deb
  to pool/main/s/schroot/dchroot_1.1.4-1_powerpc.deb
libsbuild-dev_1.1.4-1_powerpc.deb
  to pool/main/s/schroot/libsbuild-dev_1.1.4-1_powerpc.deb
libsbuild-doc_1.1.4-1_all.deb
  to pool/main/s/schroot/libsbuild-doc_1.1.4-1_all.deb
schroot-common_1.1.4-1_all.deb
  to pool/main/s/schroot/schroot-common_1.1.4-1_all.deb
schroot_1.1.4-1.diff.gz
  to pool/main/s/schroot/schroot_1.1.4-1.diff.gz
schroot_1.1.4-1.dsc
  to pool/main/s/schroot/schroot_1.1.4-1.dsc
schroot_1.1.4-1_powerpc.deb
  to pool/main/s/schroot/schroot_1.1.4-1_powerpc.deb
schroot_1.1.4.orig.tar.gz
  to pool/main/s/schroot/schroot_1.1.4.orig.tar.gz



A summary of the changes between this version and the previous one is
attached.

Thank you for reporting the bug, which will now be closed.  If you
have further comments please address them to 391319@bugs.debian.org,
and the maintainer will reopen the bug report if appropriate.

Debian distribution maintenance software
pp.
Roger Leigh <rleigh@debian.org> (supplier of updated schroot package)

(This message was generated automatically at their request; if you
believe that there is a problem with it please contact the archive
administrators by mailing ftpmaster@debian.org)


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Format: 1.7
Date: Mon, 28 May 2007 21:52:05 +0100
Source: schroot
Binary: schroot-common libsbuild-dev dchroot-dsa dchroot libsbuild-doc schroot
Architecture: source powerpc all
Version: 1.1.4-1
Distribution: unstable
Urgency: low
Maintainer: Debian buildd-tools Developers <buildd-tools-devel@lists.alioth.debian.org>
Changed-By: Roger Leigh <rleigh@debian.org>
Description: 
 dchroot    - Execute commands in a chroot environment
 dchroot-dsa - Execute commands in a chroot environment
 libsbuild-dev - development files for the Debian source builder
 libsbuild-doc - development documentation for the Debian source builder
 schroot    - Execute commands in a chroot environment
 schroot-common - common files for schroot
Closes: 391319 425964 426416
Changes: 
 schroot (1.1.4-1) unstable; urgency=low
 .
   * New upstream development release.
   * debian/control:
     - Build-Depend on the latest version of Boost, due to the change of
       library symlinks.
     - Change section of libsbuild-doc to "doc".
     - Update package descriptions.
       + Add schroot boilerplate description to schroot-common,
         libsbuild-dev and libsbuild-doc (Closes: #425964).  Thanks to
         Craig Sanders.
       + Update schroot boilerplate.
       + Add a list of reasons not to use dchroot and dchroot-dsa instead
         of schroot.
   * Processes running in the chroot on stopping a session are now killed
     by the 15killprocs setup script (Closes: #391319).  Many thanks to
     Kees Cook for implementing this.
   * The 10mount script checks if the output of schroot-listmounts is empty
     before calling umount (Closes: #426416).  Thanks to Federico Di
     Gregorio.
Files: 
 ae996c12c1a084564b2cab857c2be624 1060 admin optional schroot_1.1.4-1.dsc
 a3ddc06eb4dc18259001088a08794229 7039175 admin optional schroot_1.1.4.orig.tar.gz
 2f77bd70756346f8870a6698097e91c2 20 admin optional schroot_1.1.4-1.diff.gz
 433561c4092465b413afd84d8567708a 87614 admin optional schroot-common_1.1.4-1_all.deb
 743200cacf4cf5afa5f81453a6b59455 3245620 doc optional libsbuild-doc_1.1.4-1_all.deb
 fb3f6569374b562fc26b1e267e0ec11c 1097874 devel optional libsbuild-dev_1.1.4-1_powerpc.deb
 ecb021a6bef6fe037b034f2d70ba323e 550084 admin optional schroot_1.1.4-1_powerpc.deb
 ce1f428d3b0ee565ec3409347a7b88a1 299402 admin optional dchroot_1.1.4-1_powerpc.deb
 8315e6693596f8888985306b76896b50 299028 admin optional dchroot-dsa_1.1.4-1_powerpc.deb

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)

iD8DBQFGW1rVVcFcaSW/uEgRAqcVAJ99noHrEwaqnphr9p6H5R/DqwscawCfR7wI
zeNC6v2T7ma2xIMtDheORHo=
=6ngw
-----END PGP SIGNATURE-----

Send a report that this bug log contains spam.