]> git.proxmox.com Git - mirror_lxc.git/log
mirror_lxc.git
8 years agoFix for ALTLinux container creation in all branches
Denis Pynkin [Mon, 26 Sep 2016 17:49:47 +0000 (20:49 +0300)]
Fix for ALTLinux container creation in all branches

Use 'apt-conf' virtual package for ALTLinux default packages set

Signed-off-by: Denis Pynkin <denis_pynkin@epam.com>
8 years agoMerge pull request #1207 from brauner/2016-09-25/fix_lxc_string_split
Stéphane Graber [Sun, 25 Sep 2016 18:45:22 +0000 (14:45 -0400)]
Merge pull request #1207 from brauner/2016-09-25/fix_lxc_string_split

utils: fix lxc_string_split()

8 years agoutils: fix lxc_string_split()
Christian Brauner [Sun, 25 Sep 2016 14:51:24 +0000 (16:51 +0200)]
utils: fix lxc_string_split()

Make sure we don't return uninitialized memory.

Signed-off-by: Christian Brauner <christian.brauner@canonical.com>
8 years agoMerge pull request #1206 from cwoac/master
Christian Brauner [Sun, 25 Sep 2016 09:22:21 +0000 (11:22 +0200)]
Merge pull request #1206 from cwoac/master

Fix null derefence if attach is called without access to any tty

8 years agoFix null derefence if attach is called without access to any tty
Oliver Matthews [Sun, 25 Sep 2016 08:37:43 +0000 (09:37 +0100)]
Fix null derefence if attach is called without access to any tty

Signed-off-by: Oliver Matthews <oliver@codersoffortune.net>
8 years agoMerge pull request #1204 from tych0/close-tty-on-restore
Christian Brauner [Wed, 21 Sep 2016 22:08:14 +0000 (00:08 +0200)]
Merge pull request #1204 from tych0/close-tty-on-restore

c/r: detatch from controlling tty on restore

8 years agoc/r: detatch from controlling tty on restore
Tycho Andersen [Wed, 21 Sep 2016 21:45:49 +0000 (21:45 +0000)]
c/r: detatch from controlling tty on restore

Signed-off-by: Tycho Andersen <tycho.andersen@canonical.com>
8 years agoMerge pull request #1201 from brauner/2016-09-21/lxc_ls_no_spaces
Stéphane Graber [Wed, 21 Sep 2016 19:34:34 +0000 (15:34 -0400)]
Merge pull request #1201 from brauner/2016-09-21/lxc_ls_no_spaces

tools: do not add trailing spaces on lxc-ls -1

8 years agoMerge pull request #1202 from brauner/2016-09-21/lxc_attach_no_new_priv_fix
Stéphane Graber [Wed, 21 Sep 2016 19:33:45 +0000 (15:33 -0400)]
Merge pull request #1202 from brauner/2016-09-21/lxc_attach_no_new_priv_fix

tools: fix lxc-attach regression with -s USER

8 years agoMerge pull request #1203 from brauner/2016-09-21/retrieve_mtu_from_bridge
Stéphane Graber [Wed, 21 Sep 2016 19:32:00 +0000 (15:32 -0400)]
Merge pull request #1203 from brauner/2016-09-21/retrieve_mtu_from_bridge

2016 09 21/retrieve mtu from bridge

8 years agoconf: try to retrieve mtu from veth
Christian Brauner [Wed, 21 Sep 2016 19:07:24 +0000 (21:07 +0200)]
conf: try to retrieve mtu from veth

When the mtu cannot be retrieved from netdev->link try from veth device.

Signed-off-by: Christian Brauner <christian.brauner@canonical.com>
8 years agoconf: retrieve mtu from netdev->link
Christian Brauner [Wed, 21 Sep 2016 18:56:03 +0000 (20:56 +0200)]
conf: retrieve mtu from netdev->link

When mtu is not set, try to retrieve mtu from netdev->link.

Signed-off-by: Christian Brauner <christian.brauner@canonical.com>
8 years agotools: fix lxc-attach regression with -s USER
Christian Brauner [Wed, 21 Sep 2016 11:02:08 +0000 (13:02 +0200)]
tools: fix lxc-attach regression with -s USER

Signed-off-by: Christian Brauner <christian.brauner@canonical.com>
8 years agotools: do not add trailing spaces on lxc-ls -1
Christian Brauner [Wed, 21 Sep 2016 07:15:14 +0000 (09:15 +0200)]
tools: do not add trailing spaces on lxc-ls -1

Signed-off-by: Christian Brauner <christian.brauner@canonical.com>
8 years agoMerge pull request #1197 from ysbnim/master
Christian Brauner [Tue, 20 Sep 2016 09:35:53 +0000 (11:35 +0200)]
Merge pull request #1197 from ysbnim/master

Update Korean manuals

8 years agodoc: Add lxc.no_new_privs to Korean lxc.container.conf(5)
Sungbae Yoo [Tue, 20 Sep 2016 09:10:33 +0000 (18:10 +0900)]
doc: Add lxc.no_new_privs to Korean lxc.container.conf(5)

Update for commit 222ddc

Signed-off-by: Sungbae Yoo <sungbae.yoo@samsung.com>
8 years agoMerge pull request #1194 from tych0/cgroup-root-on-dump
Stéphane Graber [Sat, 17 Sep 2016 03:20:10 +0000 (23:20 -0400)]
Merge pull request #1194 from tych0/cgroup-root-on-dump

Cgroup root on dump

8 years agoc/r: check that cgroup_num_hierarchies > 0
Tycho Andersen [Sat, 17 Sep 2016 02:26:31 +0000 (20:26 -0600)]
c/r: check that cgroup_num_hierarchies > 0

Otherwise in the error case, we end up subtracting two from the
static_args, which would lead to a segfault :)

Signed-off-by: Tycho Andersen <tycho.andersen@canonical.com>
8 years agocgroup: drop cgroup_canonical_path
Tycho Andersen [Wed, 14 Sep 2016 14:58:38 +0000 (14:58 +0000)]
cgroup: drop cgroup_canonical_path

This is almost never the right thing to use, and we don't use it any more
anyway.

Signed-off-by: Tycho Andersen <tycho.andersen@canonical.com>
8 years agocgroup: get rid of weird hack in cgfsng_escape
Tycho Andersen [Wed, 14 Sep 2016 14:53:21 +0000 (14:53 +0000)]
cgroup: get rid of weird hack in cgfsng_escape

We initialized cgfsng in a strange way inside of its implementation of
escape so we could use it during checkpoint. Instead, the previous patch
does a hacky initialization in criu.c, and we can get rid of the hacks
elsewhere :)

Signed-off-by: Tycho Andersen <tycho.andersen@canonical.com>
8 years agoc/r: pass --cgroup-roots on checkpoint
Tycho Andersen [Wed, 14 Sep 2016 14:47:38 +0000 (14:47 +0000)]
c/r: pass --cgroup-roots on checkpoint

CRIU has added support for passing --cgroup-root on dump, which we should
use (see the criu commit 07d259f365f224b32914de26ea0fd59fc6db0001 for
details). Note that we don't have to do any version checking or anything,
because CRIU just ignored --cgroup-root on checkpoint before, so passing it
is safe, and will result in correct behavior when a sufficient version of
CRIU is present.

Signed-off-by: Tycho Andersen <tycho.andersen@canonical.com>
8 years agoutils: add lxc_deslashify
Tycho Andersen [Wed, 14 Sep 2016 14:46:47 +0000 (14:46 +0000)]
utils: add lxc_deslashify

Signed-off-by: Tycho Andersen <tycho.andersen@canonical.com>
8 years agocgroup: add new functions for interacting with hierachies
Tycho Andersen [Wed, 14 Sep 2016 14:38:46 +0000 (14:38 +0000)]
cgroup: add new functions for interacting with hierachies

N.B. that these are only implemented in cgfsng, but,

15:28:28    tych0 | do we still use cgfs anywhere? or the cgm backend?
15:29:19 stgraber | not anywhere we care about

...I think that's okay.

Signed-off-by: Tycho Andersen <tycho.andersen@canonical.com>
8 years agoc/r: fix typo in comment
Tycho Andersen [Mon, 12 Sep 2016 18:04:18 +0000 (18:04 +0000)]
c/r: fix typo in comment

Signed-off-by: Tycho Andersen <tycho.andersen@canonical.com>
8 years agoMerge pull request #1193 from hallyn/2016-09-16/thierry
Stéphane Graber [Fri, 16 Sep 2016 17:35:42 +0000 (13:35 -0400)]
Merge pull request #1193 from hallyn/2016-09-16/thierry

lxc-create -t debian fails on ppc64el arch

8 years agolxc-create -t debian fails on ppc64el arch
Thierry Fauck [Fri, 16 Sep 2016 13:34:41 +0000 (08:34 -0500)]
lxc-create -t debian fails on ppc64el arch

Template catches arch from uname -m, but for ppc64el system, arch reports ppc64le
which doesn't match image repo.

Signed-off-by: Thierry Fauck <tfauck@free.fr>
Signed-off-by: Serge Hallyn <serge@hallyn.com>
8 years agoMerge pull request #1192 from tenforward/japanese
Christian Brauner [Fri, 16 Sep 2016 07:18:45 +0000 (09:18 +0200)]
Merge pull request #1192 from tenforward/japanese

doc: Add lxc.no_new_privs to Japanese lxc.container.conf(5)

8 years agodoc: Add lxc.no_new_privs to Japanese lxc.container.conf(5)
KATOH Yasufumi [Fri, 16 Sep 2016 06:56:45 +0000 (15:56 +0900)]
doc: Add lxc.no_new_privs to Japanese lxc.container.conf(5)

Update for commit 222ddc

Signed-off-by: KATOH Yasufumi <karma@jazz.email.ne.jp>
8 years agoMerge pull request #1166 from brauner/2016-09-02/no_new_privileges
Serge Hallyn [Fri, 16 Sep 2016 01:35:21 +0000 (20:35 -0500)]
Merge pull request #1166 from brauner/2016-09-02/no_new_privileges

implement PR_SET_NO_NEW_PRIVS in liblxc

8 years agoMerge pull request #1187 from lpirl/master
Stéphane Graber [Wed, 14 Sep 2016 17:02:43 +0000 (13:02 -0400)]
Merge pull request #1187 from lpirl/master

make rsync deal with sparse files efficiently

8 years agomake rsync deal with sparse files efficiently
Lukas Pirl [Wed, 14 Sep 2016 15:40:16 +0000 (17:40 +0200)]
make rsync deal with sparse files efficiently

Signed-off-by: Lukas Pirl <git@lukas-pirl.de>
8 years agoMerge pull request #1185 from tych0/free-valid-opts-if-necessary
Christian Brauner [Tue, 13 Sep 2016 23:16:57 +0000 (01:16 +0200)]
Merge pull request #1185 from tych0/free-valid-opts-if-necessary

c/r: free valid_opts if necessary

8 years agoc/r: free valid_opts if necessary
Tycho Andersen [Tue, 13 Sep 2016 22:42:20 +0000 (16:42 -0600)]
c/r: free valid_opts if necessary

2cb80427bc468f7647309c3eca66cfc9afa85b61 introduced a malloc without a
matching free.

Signed-off-by: Tycho Andersen <tycho.andersen@canonical.com>
8 years agoMerge pull request #1184 from brauner/2016-09-13/zfs_fixes
Stéphane Graber [Tue, 13 Sep 2016 16:23:44 +0000 (12:23 -0400)]
Merge pull request #1184 from brauner/2016-09-13/zfs_fixes

lxczfs: small fixes

8 years agolxczfs: small fixes
Christian Brauner [Tue, 13 Sep 2016 15:18:23 +0000 (17:18 +0200)]
lxczfs: small fixes

- We expect destroy to fail in zfs_clone() so try to silence it so users are
  not irritated when they create zfs snapshots.
- Add -r recursive to zfs_destroy(). This code is only hit when a) the
  container has no snapshots or b) the user calls destroy with snapshots. So
  this should be safe. Without -r snapshots will remain.

Signed-off-by: Christian Brauner <christian.brauner@canonical.com>
8 years agoMerge pull request #1177 from tych0/zero-smaller-migrate-struct
Christian Brauner [Thu, 8 Sep 2016 18:52:52 +0000 (20:52 +0200)]
Merge pull request #1177 from tych0/zero-smaller-migrate-struct

c/r: zero a smaller than known migrate_opts struct

8 years agoc/r: zero a smaller than known migrate_opts struct
Tycho Andersen [Thu, 8 Sep 2016 15:14:42 +0000 (09:14 -0600)]
c/r: zero a smaller than known migrate_opts struct

Signed-off-by: Tycho Andersen <tycho.andersen@canonical.com>
8 years agoMerge pull request #1173 from melato/alpine-cron-v2
Christian Brauner [Tue, 6 Sep 2016 14:15:44 +0000 (16:15 +0200)]
Merge pull request #1173 from melato/alpine-cron-v2

templates: use correct cron version in alpine template

8 years agotemplates: use correct cron version in alpine template
Alex Athanasopoulos [Tue, 6 Sep 2016 14:09:55 +0000 (17:09 +0300)]
templates: use correct cron version in alpine template

Signed-off-by: Alex Athanasopoulos <alex@melato.org>
8 years agotests: add test for PR_SET_NO_NEW_PRIVS
Christian Brauner [Sat, 3 Sep 2016 11:59:47 +0000 (13:59 +0200)]
tests: add test for PR_SET_NO_NEW_PRIVS

Signed-off-by: Christian Brauner <christian.brauner@canonical.com>
8 years agoattach, start: declare PR_{S,G}PR_GET_NO_NEW_PRIVS
Christian Brauner [Sat, 3 Sep 2016 13:19:27 +0000 (15:19 +0200)]
attach, start: declare PR_{S,G}PR_GET_NO_NEW_PRIVS

Signed-off-by: Christian Brauner <christian.brauner@canonical.com>
8 years agodoc: add lxc.no_new_privs to lxc.container.conf
Christian Brauner [Sat, 3 Sep 2016 06:00:20 +0000 (08:00 +0200)]
doc: add lxc.no_new_privs to lxc.container.conf

Signed-off-by: Christian Brauner <christian.brauner@canonical.com>
8 years agoattach: use PR_SET_NO_NEW_PRIVS
Christian Brauner [Fri, 2 Sep 2016 16:56:48 +0000 (18:56 +0200)]
attach: use PR_SET_NO_NEW_PRIVS

- When we detect that the container, we want to attach to, has been stared with
  PR_SET_NO_NEW_PRIVS we attach with PR_SET_NO_NEW_PRIVS as well. (We might
  relax this restriction later but let's be strict for now.)
- When LXC_ATTACH_NO_NEW_PRIVS is set in the flags passed to
  lxc_attach()/attach_child_main() then we set PR_SET_NO_NEW_PRIVS irrespective
  of whether the container was started with PR_SET_NO_NEW_PRIVS or not.
- Set no_new_privs before lsm and seccomp. We probably don't want attach() to
  be able to change the lsm or seccomp policy if the container was started with
  PR_SET_NO_NEW_PRIVS enabled.

Signed-off-by: Christian Brauner <christian.brauner@canonical.com>
8 years agoattach: call lxc_container_new() earlier
Christian Brauner [Fri, 2 Sep 2016 16:39:11 +0000 (18:39 +0200)]
attach: call lxc_container_new() earlier

We will reuse the newly initialized container for PR_SET_NO_NEW_PRIVS.

Signed-off-by: Christian Brauner <christian.brauner@canonical.com>
8 years agoattach_options: add LXC_ATTACH_NO_NEW_PRIVS
Christian Brauner [Fri, 2 Sep 2016 16:17:11 +0000 (18:17 +0200)]
attach_options: add LXC_ATTACH_NO_NEW_PRIVS

Add a flag for PR_SET_NO_NEW_PRIVS. It is off by default.

Signed-off-by: Christian Brauner <christian.brauner@canonical.com>
8 years agostart: set PR_SET_NO_NEW_PRIVS when requested
Christian Brauner [Thu, 1 Sep 2016 23:40:39 +0000 (01:40 +0200)]
start: set PR_SET_NO_NEW_PRIVS when requested

Set no_new_privs after setting the lsm label. If we do set it before we aren't
allowed to change the label anymore.

Signed-off-by: Christian Brauner <christian.brauner@canonical.com>
8 years agoconf, confile: add option for PR_SET_NO_NEW_PRIVS
Christian Brauner [Thu, 1 Sep 2016 23:30:59 +0000 (01:30 +0200)]
conf, confile: add option for PR_SET_NO_NEW_PRIVS

Signed-off-by: Christian Brauner <christian.brauner@canonical.com>
8 years agoMerge pull request #1168 from jerome-pouiller/master
Christian Brauner [Mon, 5 Sep 2016 18:01:18 +0000 (20:01 +0200)]
Merge pull request #1168 from jerome-pouiller/master

configure: add --disable-werror

8 years agoMerge pull request #1169 from brauner/2016-09-04/fix_syslog
Stéphane Graber [Mon, 5 Sep 2016 16:51:57 +0000 (12:51 -0400)]
Merge pull request #1169 from brauner/2016-09-04/fix_syslog

syslog tweaks

8 years agoMerge pull request #1167 from brauner/2016-09-03/fix_log_name
Stéphane Graber [Mon, 5 Sep 2016 16:48:25 +0000 (12:48 -0400)]
Merge pull request #1167 from brauner/2016-09-03/fix_log_name

console: use correct log name

8 years agotests: add lxc.syslog tests to get_item
Christian Brauner [Sun, 4 Sep 2016 22:24:24 +0000 (00:24 +0200)]
tests: add lxc.syslog tests to get_item

Signed-off-by: Christian Brauner <christian.brauner@mailbox.org>
8 years agotests: fix get_item tests
Christian Brauner [Sun, 4 Sep 2016 22:01:51 +0000 (00:01 +0200)]
tests: fix get_item tests

Signed-off-by: Christian Brauner <christian.brauner@mailbox.org>
8 years agosyslog: simplify and model after lxc log functions
Christian Brauner [Sun, 4 Sep 2016 19:16:12 +0000 (21:16 +0200)]
syslog: simplify and model after lxc log functions

- add lxc_syslog_priority_to_string()
- add lxc_syslog_priority_to_int()
- remove syslog_facility struct
- add lxc.syslog to lxc_getconfig struct
- adapt config_syslog() callback

Signed-off-by: Christian Brauner <christian.brauner@mailbox.org>
8 years agoconfigure: add --disable-werror
Jérôme Pouiller [Sun, 4 Sep 2016 07:35:44 +0000 (09:35 +0200)]
configure: add --disable-werror

-Werror may break builds on some scenarios with trivialities
(especially during developments).

Signed-off-by: Jérôme Pouiller <jezz@sysmic.org>
8 years agoconsole: use correct log name
Christian Brauner [Sat, 3 Sep 2016 18:05:54 +0000 (20:05 +0200)]
console: use correct log name

lxc_console is used with lxc_console.c

Signed-off-by: Christian Brauner <christian.brauner@mailbox.org>
8 years agoMerge pull request #1163 from mabes/fix-libetc-dir
Christian Brauner [Thu, 1 Sep 2016 15:00:16 +0000 (17:00 +0200)]
Merge pull request #1163 from mabes/fix-libetc-dir

templates: remove creation of bogus directory in Debian templates

8 years agotemplates: remove creation of bogus directory
Maxime BESSON [Thu, 1 Sep 2016 14:09:37 +0000 (16:09 +0200)]
templates: remove creation of bogus directory

An incorrect quoting introduced in bf39edb caused a /{lib,etc} folder to
appear in Debian templates

The very next line :
    mkdir -p "${rootfs}/etc/systemd/system/getty.target.wants

makes creating ${rootfs}/etc/systemd/system/ unnecessary in the first
place

Signed-off-by: Maxime Besson <maxime.besson@smile.fr>
8 years agoMerge pull request #1162 from brauner/2016-08-31/remove_halt_symlink
Stéphane Graber [Wed, 31 Aug 2016 22:29:32 +0000 (18:29 -0400)]
Merge pull request #1162 from brauner/2016-08-31/remove_halt_symlink

templates: rm halt.target -> sigpwr.target symlink

8 years agotemplates: rm halt.target -> sigpwr.target symlink
Christian Brauner [Wed, 31 Aug 2016 21:53:50 +0000 (23:53 +0200)]
templates: rm halt.target -> sigpwr.target symlink

Given commit 330ae3d350e060e5702a0e5ef5d0faeeeea8df6e:

    lxccontainer: detect if we should send SIGRTMIN+3

    This is required by systemd to cleanly shutdown. Other init systems should not
    have SIGRTMIN+3 in the blocked signals set.

we should stop symlinking halt.target to sigpwr.target for systemd.

Signed-off-by: Christian Brauner <cbrauner@suse.de>
8 years agoMerge pull request #1157 from evgeni/doxygen-no-full-path-names
Christian Brauner [Sat, 27 Aug 2016 10:21:29 +0000 (12:21 +0200)]
Merge pull request #1157 from evgeni/doxygen-no-full-path-names

set FULL_PATH_NAMES=NO in doc/api/Doxyfile

8 years agoset FULL_PATH_NAMES=NO in doc/api/Doxyfile
Evgeni Golov [Sat, 27 Aug 2016 10:14:53 +0000 (12:14 +0200)]
set FULL_PATH_NAMES=NO in doc/api/Doxyfile

otherwise the generated docs have the full build path in them
and nonbody cares that the files were built in
 /build/lxc-_BVY2u/lxc-2.0.4/src/lxc/

Signed-off-by: Evgeni Golov <evgeni@debian.org>
8 years agoMerge pull request #1155 from tych0/migration-fixes
Christian Brauner [Fri, 26 Aug 2016 21:53:35 +0000 (23:53 +0200)]
Merge pull request #1155 from tych0/migration-fixes

Migration fixes

8 years agoc/r: write status only after trying to parse the pid
Tycho Andersen [Fri, 26 Aug 2016 20:29:45 +0000 (16:29 -0400)]
c/r: write status only after trying to parse the pid

Previously, we write a "success" status but tried to parse the pid. This
meant that we wouldn't notice a successful restore but failure to parse the
pid, which was a little strange.

We still don't know the child pid, so we will end up with a restored
process tree and a running container, but at least in this case the API
will return false indicating that something failed.

We could kill(-1, 9) in this case, but since liblxc runs as root sometimes
(e.g. LXD), that would be a Very Bad Thing.

Signed-off-by: Tycho Andersen <tycho.andersen@canonical.com>
8 years agoremove extra 'ret'
Tycho Andersen [Fri, 26 Aug 2016 20:13:06 +0000 (16:13 -0400)]
remove extra 'ret'

Signed-off-by: Tycho Andersen <tycho.andersen@canonical.com>
8 years agoMerge pull request #1154 from stgraber/master
Christian Brauner [Fri, 26 Aug 2016 20:07:58 +0000 (22:07 +0200)]
Merge pull request #1154 from stgraber/master

c/r: Fix pid_t on some arches

8 years agoc/r: Fix pid_t on some arches
Stéphane Graber [Fri, 26 Aug 2016 19:41:34 +0000 (15:41 -0400)]
c/r: Fix pid_t on some arches

Signed-off-by: Stéphane Graber <stgraber@ubuntu.com>
8 years agoMerge pull request #1152 from tych0/fix-migration-2.0.4
Christian Brauner [Fri, 26 Aug 2016 16:48:38 +0000 (18:48 +0200)]
Merge pull request #1152 from tych0/fix-migration-2.0.4

c/r: use /proc/self/tid/children instead of pidfile

8 years agoc/r: use /proc/self/tid/children instead of pidfile
Tycho Andersen [Fri, 26 Aug 2016 16:07:19 +0000 (16:07 +0000)]
c/r: use /proc/self/tid/children instead of pidfile

All we really needed a unique temp file for was passing the pid. Since CRIU
opened this with O_EXCL | O_CREAT, this was "safe" (users could still
overwrite it afterwards, but the monitor would immediately die since the
only valid number in there was the init process).

In any case, we can just read /proc/self/tid/children, which lists the
child process.

Closes #1150

Signed-off-by: Tycho Andersen <tycho.andersen@canonical.com>
8 years agoMerge pull request #1151 from brauner/2016-08-26/fix_lxc_device
Serge Hallyn [Fri, 26 Aug 2016 14:48:29 +0000 (10:48 -0400)]
Merge pull request #1151 from brauner/2016-08-26/fix_lxc_device

tools: use boolean for ret in lxc_device.c

8 years agotools: use boolean for ret in lxc_device.c
Christian Brauner [Fri, 26 Aug 2016 10:49:23 +0000 (12:49 +0200)]
tools: use boolean for ret in lxc_device.c

Signed-off-by: Christian Brauner <cbrauner@suse.de>
8 years agoMerge pull request #1139 from brauner/2016-08-18/use_exit_in_all_tools
Serge Hallyn [Fri, 26 Aug 2016 10:15:37 +0000 (06:15 -0400)]
Merge pull request #1139 from brauner/2016-08-18/use_exit_in_all_tools

tools: use exit(EXIT_*) everywhere

8 years agoMerge pull request #1145 from glensc/pld-patch
Christian Brauner [Tue, 23 Aug 2016 19:33:10 +0000 (21:33 +0200)]
Merge pull request #1145 from glensc/pld-patch

add PLD Linux support to upstream repo

8 years agotemplates: add PLD Linux support
Elan Ruusamäe [Sat, 20 Aug 2016 13:51:04 +0000 (16:51 +0300)]
templates: add PLD Linux support

from pld-linux repo:
https://github.com/pld-linux/lxc/commits/auto/th/lxc-2.0.4-1/lxc-pld.in.sh
https://github.com/pld-linux/lxc/commits/auto/th/lxc-2.0.4-1/lxc-pld.patch

Signed-off-by: Elan Ruusamäe <glen@delfi.ee>
8 years agoMerge pull request #1148 from glensc/fed-openssl
Christian Brauner [Tue, 23 Aug 2016 16:58:08 +0000 (18:58 +0200)]
Merge pull request #1148 from glensc/fed-openssl

templates: fedora requires openssl binary

8 years agotemplates: fedora requires openssl binary
Elan Ruusamäe [Tue, 23 Aug 2016 16:31:38 +0000 (19:31 +0300)]
templates: fedora requires openssl binary

/usr/share/lxc/templates/lxc-fedora: line 1078: openssl: command not found

Signed-off-by: Elan Ruusamäe <glen@delfi.ee>
8 years agoMerge pull request #1146 from glensc/fd-200
Christian Brauner [Mon, 22 Aug 2016 07:26:59 +0000 (09:26 +0200)]
Merge pull request #1146 from glensc/fd-200

use fd 9 instead of 200

8 years agotemplates: use fd 9 instead of 200
Elan Ruusamäe [Sat, 20 Aug 2016 14:09:28 +0000 (17:09 +0300)]
templates: use fd 9 instead of 200

to catch up mksh changes from 17abf27

Signed-off-by: Elan Ruusamäe <glen@delfi.ee>
8 years agoMerge pull request #1144 from brauner/2016-08-19/seccomp_non_functional_changes
Stéphane Graber [Fri, 19 Aug 2016 20:05:14 +0000 (16:05 -0400)]
Merge pull request #1144 from brauner/2016-08-19/seccomp_non_functional_changes

seccomp: non functional changes

8 years agoseccomp: non functional changes
Christian Brauner [Fri, 19 Aug 2016 16:53:02 +0000 (18:53 +0200)]
seccomp: non functional changes

- log more errnos
- adapt coding style

Signed-off-by: Christian Brauner <cbrauner@suse.de>
8 years agoMerge pull request #1141 from brauner/2016-08-19/add_missing_arches
Stéphane Graber [Fri, 19 Aug 2016 15:58:04 +0000 (11:58 -0400)]
Merge pull request #1141 from brauner/2016-08-19/add_missing_arches

confile: add more archs to lxc_config_parse_arch()

8 years agoMerge pull request #1142 from brauner/2016-08-19/add_s390x_seccomp_support
Stéphane Graber [Fri, 19 Aug 2016 15:57:28 +0000 (11:57 -0400)]
Merge pull request #1142 from brauner/2016-08-19/add_s390x_seccomp_support

seccomp: add support for s390x

8 years agoseccomp: remove double include and order includes
Christian Brauner [Fri, 19 Aug 2016 13:20:22 +0000 (15:20 +0200)]
seccomp: remove double include and order includes

Signed-off-by: Christian Brauner <cbrauner@suse.de>
8 years agoseccomp: add support for s390x
Christian Brauner [Fri, 19 Aug 2016 09:57:14 +0000 (11:57 +0200)]
seccomp: add support for s390x

Signed-off-by: Christian Brauner <cbrauner@suse.de>
8 years agoconfile: add more archs to lxc_config_parse_arch()
Christian Brauner [Fri, 19 Aug 2016 09:15:30 +0000 (11:15 +0200)]
confile: add more archs to lxc_config_parse_arch()

Signed-off-by: Christian Brauner <cbrauner@suse.de>
8 years agoMerge pull request #1133 from jcowgill/mips
Christian Brauner [Thu, 18 Aug 2016 16:49:54 +0000 (18:49 +0200)]
Merge pull request #1133 from jcowgill/mips

Add MIPS support

8 years agoseccomp: fix strerror()
James Cowgill [Thu, 18 Aug 2016 15:48:24 +0000 (16:48 +0100)]
seccomp: fix strerror()

Signed-off-by: James Cowgill <james410@cowgill.org.uk>
8 years agoseccomp: Add mips and mips64 entries to lxc_config_parse_arch
James Cowgill [Thu, 18 Aug 2016 13:27:35 +0000 (14:27 +0100)]
seccomp: Add mips and mips64 entries to lxc_config_parse_arch

Fixes "unsupported personality" warnings when starting containers.

Signed-off-by: James Cowgill <james410@cowgill.org.uk>
8 years agotemplates: Add mips hostarch detection to debian
James Cowgill [Mon, 15 Aug 2016 16:10:00 +0000 (16:10 +0000)]
templates: Add mips hostarch detection to debian

Signed-off-by: James Cowgill <james410@cowgill.org.uk>
8 years agoseccomp: Implement MIPS seccomp handling
James Cowgill [Mon, 15 Aug 2016 16:09:44 +0000 (16:09 +0000)]
seccomp: Implement MIPS seccomp handling

MIPS processors implement 3 ABIs: o32, n64 and n32 (similar to x32). The kernel
treats each ABI separately so syscalls disallowed on "all" arches should be
added to all three seccomp sets. This is implemented by expanding compat_arch
and compat_ctx to accept two compat architectures.

After this, the MIPS hostarch detection code and config section code is added.

Signed-off-by: James Cowgill <james410@cowgill.org.uk>
8 years agoutils: Add mips signalfd syscall numbers
James Cowgill [Fri, 12 Aug 2016 15:54:14 +0000 (15:54 +0000)]
utils: Add mips signalfd syscall numbers

Signed-off-by: James Cowgill <james410@cowgill.org.uk>
8 years agotools: unify exit() calls outside of main()
Christian Brauner [Thu, 18 Aug 2016 09:11:34 +0000 (11:11 +0200)]
tools: unify exit() calls outside of main()

This also changes the help exit status of lxc-unshare and lxc-usernsexec. So
far they did exit(EXIT_FAILURE) whereas all other tools do exit(EXIT_SUCCESS).
Let's align them with the rest of the tools. This should be safe to do.

Signed-off-by: Christian Brauner <cbrauner@suse.de>
8 years agotools: use exit(EXIT_*) everywhere
Christian Brauner [Thu, 18 Aug 2016 07:16:41 +0000 (09:16 +0200)]
tools: use exit(EXIT_*) everywhere

Signed-off-by: Christian Brauner <cbrauner@suse.de>
8 years agoMerge pull request #1140 from ysbnim/master
Christian Brauner [Thu, 18 Aug 2016 08:55:18 +0000 (10:55 +0200)]
Merge pull request #1140 from ysbnim/master

Update korean manuals

8 years agodoc: Add --rcfile to Japanese common opts
Sungbae Yoo [Thu, 18 Aug 2016 08:09:00 +0000 (17:09 +0900)]
doc: Add --rcfile to Japanese common opts

Update for commit 71d74a8

Signed-off-by: Sungbae Yoo <sungbae.yoo@samsung.com>
8 years agodoc: Add --rcfile to Korean common opts
Sungbae Yoo [Thu, 18 Aug 2016 07:49:26 +0000 (16:49 +0900)]
doc: Add --rcfile to Korean common opts

Update for commit 71d74a8

Signed-off-by: Sungbae Yoo <sungbae.yoo@samsung.com>
8 years agodoc: Update Korean lxc-attach(1)
Sungbae Yoo [Thu, 18 Aug 2016 07:36:24 +0000 (16:36 +0900)]
doc: Update Korean lxc-attach(1)

 * Update for commit 03b0398
 * Fix mistranslation words

Signed-off-by: Sungbae Yoo <sungbae.yoo@samsung.com>
8 years agoMerge pull request #1138 from stgraber/master
Christian Brauner [Wed, 17 Aug 2016 19:49:30 +0000 (21:49 +0200)]
Merge pull request #1138 from stgraber/master

Use full GPG fingerprint instead of long IDs.

8 years agoUse full GPG fingerprint instead of long IDs.
Stéphane Graber [Wed, 17 Aug 2016 19:42:34 +0000 (15:42 -0400)]
Use full GPG fingerprint instead of long IDs.

With how easy it is to create a collision on a short ID nowadays and
given that the user doesn't actually have to remember or manually enter
the key ID, lets just use the full fingerprint from now on.

Signed-off-by: Stéphane Graber <stgraber@ubuntu.com>
8 years agoMerge pull request #1122 from Blub/rcfile-as-common-arg
Christian Brauner [Wed, 17 Aug 2016 15:27:16 +0000 (17:27 +0200)]
Merge pull request #1122 from Blub/rcfile-as-common-arg

Rcfile as common arg

8 years agocleanup: replace tabs wth spaces in usage strings
Wolfgang Bumiller [Fri, 12 Aug 2016 10:33:44 +0000 (12:33 +0200)]
cleanup: replace tabs wth spaces in usage strings

Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>