Clear env before starting a container. Do it right before setting
the container=lxc variable.
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Acked-by: Stéphane Graber <stgraber@ubuntu.com>
(I'll be out until Jan 2, but in the meantime, here is hopefully a
little newyears gift - this seems to allow lxc-start with / being
MS_SHARED on the host)
When / is MS_SHARED (for instance with f18 and modern arch), lxc-start
fails on pivot_root. The kernel enforces that, when doing pivot_root,
the parent of current->fs->root (as well as the new root and the putold
location) not be MS_SHARED.
To work around this, check /proc/self/mountinfo for a 'shared:' in
the '/' line. If it is there, then create a tiny MS_SLAVE tmpfs dir to
serve as parent of /, recursively bind mount / into /root under that dir,
make it rslave, and chroot into it.
Tested with ubuntu raring image after doing 'mount --make-rshared /'.
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Acked-by: Stéphane Graber <stgraber@ubuntu.com>
Currently, lxc-destory will attempt to destroy a container if it is not in
the RUNNING state, but doing so is not good when the container is FROZEN, or
in other transitional states.
Signed-off-by: Dwight Engen <dwight.engen@oracle.com>
Acked-by: Serge E. Hallyn <serge.hallyn@ubuntu.com>
If you start more than one lxc-start/lxc-execute with the same name at the
same time, or just do an lxc-start/lxc-execute with the name of a container
that is already running, lxc doesn't figure out that the container with this
name is already running until fairly late in the initialization process: ie
when __lxc_start() -> lxc_poll() -> lxc_command_mainloop_add() attempts to
create the same abstract socket name.
By this point a fair amount of initialization has been done that actually
messes up the running container. For example __lxc_start() -> lxc_spawn() ->
lxc_cgroup_create() -> lxc_one_cgroup_create() -> try_to_move_cgname() moves
the running container's cgroup to a name of deadXXXXXX.
The solution in this patch is to use the atomic existence of the abstract
socket name as the indicator that the container is already running. To do
so, I just refactored lxc_command_mainloop_add() into an lxc_command_init()
routine that attempts to bind the socket, and ensure this is called earlier
before much initialization has been done.
In testing, I verified that maincmd_fd was still open at the time of lxc_fini,
so the entire lifetime of the container's run should be covered. The only
explicit close of this fd was in the reboot case of lxcapi_start(), which is
now moved to lxc_fini(), which I think is more appropriate.
Even though it is not checked any more, set maincmd_fd to -1 instead of 0 to
indicate its not open since 0 could be a valid fd.
Signed-off-by: Dwight Engen <dwight.engen@oracle.com>
Acked-by: Serge E. Hallyn <serge.hallyn@ubuntu.com>
For example doing "lxc-execute -n tmpct /bin/bash" will call setup_kmsg(), but
in this case rootfs->mount/dev directory doesn't even exist so the call to
symlink fails with ENOENT. Commit f62b3449 made this failure not fatal, but
we should not even try it when we know it will fail. See similar code in
setup_tty(), setup_console(), etc.
Signed-off-by: Dwight Engen <dwight.engen@oracle.com>
Acked-by: Serge E. Hallyn <serge.hallyn@ubuntu.com>
Acked-by: Stéphane Graber <stgraber@ubuntu.com>
Update add_device_node to use the new set_cgroup_item call instead
of having to figure out the cgroup paths and update the entries manually.
Signed-off-by: Stéphane Graber <stgraber@ubuntu.com>
Acked-by: Serge E. Hallyn <serge.hallyn@ubuntu.com>
Updates the binding for the two new functions.
This also fixes some problems with the argument checking of
get_config_item that'd otherwise lead to a segfault.
The python binding for set_cgroup_item and get_cgroup_item are pretty
raw as lxc has little control over the cgroup entries.
That means that we don't try to interpret lists as we do for the config
entries.
Signed-off-by: Stéphane Graber <stgraber@ubuntu.com>
Acked-by: Serge E. Hallyn <serge.hallyn@ubuntu.com>
The python binding was forcing the user to pass a base path to
get_keys() even though the C binding doesn't require it.
Signed-off-by: Stéphane Graber <stgraber@ubuntu.com>
Acked-by: Serge E. Hallyn <serge.hallyn@ubuntu.com>
set_cgroup_item takes a pointer to a running container, a cgroup subsystem
name, and a char *value and it mimicks
'lxc-cgroup -n containername subsys value'
get_cgroup_item takes a pointer to a running container, a a cgroup
subsystem name, a destination value * and the length of the value being
sent in, and returns the length of what was read from the cgroup file.
If a 0 len is passed in, then the length of the file is returned. So
you can do
len = c->get_cgroup_item(c, "devices.list", NULL, 0);
v = malloc(len+1);
ret = c->get_cgroup_item(c, "devices.list", v, len);
to read the whole file.
This patch also disables the lxc-init part of the startone test, which
was failing because lxc-init has been moved due to multiarch issues.
The test is salvagable, but saving it was beyond this effort.
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Acked-by: Stéphane Graber <stgraber@ubuntu.com>
This restores an old behaviour where lxc-create can be called without
a template. In such case, only a minimal configuration is built and no
rootfs is created. However the various backingstore code is still used.
Signed-off-by: Stéphane Graber <stgraber@ubuntu.com>
Acked-by: Serge E. Hallyn <serge.hallyn@ubuntu.com>
Switch the python scripts to using @LXCPATH@.
According to grep, this was the last occurence of a /var/*/lxc
path in the code.
Signed-off-by: Stéphane Graber <stgraber@ubuntu.com>
Acked-by: Serge E. Hallyn <serge.hallyn@ubuntu.com>
Move to per-template lock (except for oracle that's per-container).
Also ensure that the path used for the lock is relative to LOCALSTATEDIR.
Signed-off-by: Stéphane Graber <stgraber@ubuntu.com>
Acked-by: Serge E. Hallyn <serge.hallyn@ubuntu.com>
Use LOCALSTATEDIR to generate the path to the cache.
Signed-off-by: Stéphane Graber <stgraber@ubuntu.com>
Acked-by: Serge E. Hallyn <serge.hallyn@ubuntu.com>
This commit updates all scripts using mixed indent to a consistent
4 spaces indent.
In the past quite a few of those scripts used tabs to instead of 8 spaces or
instead of 4 spaces, sometimes mixing those in the same line and sometimes
changing the tab width within the same file.
Signed-off-by: Stéphane Graber <stgraber@ubuntu.com>
Acked-by: Serge E. Hallyn <serge.hallyn@ubuntu.com>
This updates all the templates and the configuration files to consistently
use "key = value" everywhere.
Signed-off-by: Stéphane Graber <stgraber@ubuntu.com>
Acked-by: Serge E. Hallyn <serge.hallyn@ubuntu.com>
- Update COPYING to the current copy of the LPGL-2.1 license from
common-licences (only difference is some indentation).
- Remove mixed tabs/spaces in CONTRIBUTING
- Make INSTALL fit on 79 cols.
Signed-off-by: Stéphane Graber <stgraber@ubuntu.com>
Acked-by: Serge E. Hallyn <serge.hallyn@ubuntu.com>
This adds a section about the staging branch to the README and updates
the list of supported architectures.
Signed-off-by: Stéphane Graber <stgraber@ubuntu.com>
Acked-by: Serge E. Hallyn <serge.hallyn@ubuntu.com>
Let oracle template work when host is fedora or oracle and the lsb_release
command is not present. Verify the arch given is valid. Don't add lxc.network
section again if already present.
Signed-off-by: Dwight Engen <dwight.engen@oracle.com>
Acked-by: Stéphane Graber <stgraber@ubuntu.com>
lxc-archlinux was apparently left out of the last change, apply
the same modification as the other templates.
Signed-off-by: Stéphane Graber <stgraber@ubuntu.com>
Re-arrange the code so that we only grab the container object when doing
something more than building a simple list of existing containers.
This means that now the following calls can run unprivileged:
- lxc-ls
- lxc-ls -1
Everything else will still require root privileges.
Signed-off-by: Stéphane Graber <stgraber@ubuntu.com>
Acked-by: Serge E. Hallyn <serge.hallyn@ubuntu.com>
This commit does the following changes to the python API:
- Rename the add_device API call to add_device_node
- Adds an extra check that the container is running to add_device_node
- Introduces a new add_device_net function
And the following changes to the lxc-device tool:
- Change parser setup to better cope with variable number of arguments
- Add support for network devices (currently auto-detected)
- Support for different names on the host and in the container
Signed-off-by: Stéphane Graber <stgraber@ubuntu.com>
Acked-by: Serge E. Hallyn <serge.hallyn@ubuntu.com>
Instead of returning a python stacktrace, check what the current euid is
and show an argparse error message similar to that used in lxc-start-ephemeral.
Signed-off-by: Stéphane Graber <stgraber@ubuntu.com>
Acked-by: Serge E. Hallyn <serge.hallyn@ubuntu.com>
Instead of returning a python stacktrace, check what the current euid is
and show an argparse error message similar to that used in lxc-start-ephemeral.
Signed-off-by: Stéphane Graber <stgraber@ubuntu.com>
Acked-by: Serge E. Hallyn <serge.hallyn@ubuntu.com>
Use argparse's error function instead of our own print + exit.
Signed-off-by: Stéphane Graber <stgraber@ubuntu.com>
Acked-by: Serge E. Hallyn <serge.hallyn@ubuntu.com>
When a physical nic is being set up, store its ifindex and original name
in struct lxc_conf. At reboot, reset the original name.
We can't just go over the original network list in lxc_conf at shutdown
because that may be tweaked in the meantime through the C api. The
saved_nics list is only setup during lxc_spawn(), and restored and
freed after lxc_start.
Bug-Ubuntu: https://bugs.launchpad.net/ubuntu/+source/lxc/+bug/1086244
Changelog: remove non-effect change in execute.c
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Acked-by: Stéphane Graber <stgraber@ubuntu.com>
The $LXCPATH (default /var/lib/lxc) directory was not being created by
make install, so unless it gets created by some other means
(packaging tools), commands such as lxc-create will fail.
Signed-off-by: Dwight Engen <dwight.engen@oracle.com>
Acked-by: Stéphane Graber <stgraber@ubuntu.com>
Re-introduce the old lxc-ls script and manpage under a new legacy
sub-directory.
Those will be installed in place of their python equivalent when LXC
is built without --enable-python.
Any other script ported to python should be added to those lists.
Signed-off-by: Stéphane Graber <stgraber@ubuntu.com>
Acked-by: Serge E. Hallyn <serge.hallyn@ubuntu.com>
This rewrite is mostly compatible with the shell version.
--active and -1 still work and behave as they used to.
This adds --running, --stopped and --frozen as state filters.
A new "fancy" view is also implemented (can be used with --fancy) and
will show containers in a column-based interface with the following fields:
- name
- state
- ipv4
- ipv6
- pid of init
Signed-off-by: Stéphane Graber <stgraber@ubuntu.com>
Acked-by: Serge E. Hallyn <serge.hallyn@ubuntu.com>
Add 'lxc.logfile' and 'lxc.loglevel' config items. Values provided on
the command line override the config items.
Have lxccontainer not set a default loglevel and logfile.
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Acked-by: Stéphane Graber <stgraber@ubuntu.com>
I'm not sure whether we want this: is -H ubiquitous?
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Acked-by: Michael H. Warfield <mhw@WittsEnd.com>
Acked-by: Dwight Engen <dwight.engen@oracle.com>
Otherwise busybox clones have been reported to explode from
2M to 440M.
Reported-by: Rene K. Mueller <spiritdude@gmail.com>
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Acked-by: Michael H. Warfield <mhw@WittsEnd.com>
Acked-by: Dwight Engen <dwight.engen@oracle.com>
