[lxc-devel] [PATCH v3 6/6] lxc-attach: Add -R option to remount /sys and /proc when only partially attaching
Serge Hallyn
serge.hallyn at canonical.com
Thu May 24 14:28:55 UTC 2012
Quoting Christian Seiler (christian at iwakd.de):
> When attaching to only some namespaces of the container but not the mount
> namespace, the contents of /sys and /proc of the host system do not properly
> reflect the context of the container's pid and/or network namespaces, and
> possibly others.
>
> The introduced -R option adds the possibility to additionally unshare the
> mount namespace (when it is not being attached) and remount /sys and /proc
> in order for those filesystems to properly reflect the container's context
> even when only attaching to some of the namespaces.
>
> Signed-off-by: Christian Seiler <christian at iwakd.de>
> Acked-by: Serge Hallyn <serge.hallyn at canonical.com>
Note there is no reason to resend this patch for this, but do you think
it would be worthwhile to warn if the user specified -R, but CLONE_NEWNS
was already in the mount flags?
> Cc: Daniel Lezcano <daniel.lezcano at free.fr>
> ---
> doc/lxc-attach.sgml.in | 44 +++++++++++++++++++++++++++++++++++---------
> src/lxc/attach.c | 44 ++++++++++++++++++++++++++++++++++++++++++++
> src/lxc/attach.h | 1 +
> src/lxc/lxc_attach.c | 22 +++++++++++++++++++++-
> 4 files changed, 101 insertions(+), 10 deletions(-)
>
> diff --git a/doc/lxc-attach.sgml.in b/doc/lxc-attach.sgml.in
> index 035cd27..1724393 100644
> --- a/doc/lxc-attach.sgml.in
> +++ b/doc/lxc-attach.sgml.in
> @@ -50,7 +50,7 @@ Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
> <cmdsynopsis><command>lxc-attach <replaceable>-n
> name</replaceable> <optional>-a
> arch</optional> <optional>-e</optional> <optional>-s
> - namespaces</optional>
> + namespaces</optional> <optional>-R</optional>
> <optional>-- command</optional></command></cmdsynopsis>
> </refsynopsisdiv>
>
> @@ -146,7 +146,30 @@ Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
> </listitem>
> </varlistentry>
>
> - </variablelist>
> + <varlistentry>
> + <term>
> + <option>-R, --remount-sys-proc</option>
> + </term>
> + <listitem>
> + <para>
> + When using <option>-s</option> and the mount namespace is not
> + included, this flag will cause <command>lxc-attach</command>
> + to remount <replaceable>/proc</replaceable> and
> + <replaceable>/sys</replaceable> to reflect the current other
> + namespace contexts.
> + </para>
> + <para>
> + Please see the <emphasis>Notes</emphasis> section for more
> + details.
> + </para>
> + <para>
> + This option will be ignored if one tries to attach to the
> + mount namespace anyway.
> + </para>
> + </listitem>
> + </varlistentry>
> +
> + </variablelist>
>
> </refsect1>
>
> @@ -229,13 +252,16 @@ Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
> the network namespace.
> </para>
> <para>
> - A workaround is to use <command>lxc-unshare</command> to unshare
> - the mount namespace after using <command>lxc-attach</command> with
> - <replaceable>-s PID</replaceable> and/or <replaceable>-s
> - NETWORK</replaceable> and then unmount and then mount again both
> - pseudo-filesystems within that new mount namespace, before
> - executing a program/script that relies on this information to be
> - correct.
> + To work around this problem, the <option>-R</option> flag provides
> + the option to remount <replaceable>/proc</replaceable> and
> + <replaceable>/sys</replaceable> in order for them to reflect the
> + network/pid namespace context of the attached process. In order
> + not to interfere with the host's actual filesystem, the mount
> + namespace will be unshared (like <command>lxc-unshare</command>
> + does) before this is done, esentially giving the process a new
> + mount namespace, which is identical to the hosts's mount namespace
> + except for the <replaceable>/proc</replaceable> and
> + <replaceable>/sys</replaceable> filesystems.
> </para>
> </refsect1>
>
> diff --git a/src/lxc/attach.c b/src/lxc/attach.c
> index 37e667f..ec0e083 100644
> --- a/src/lxc/attach.c
> +++ b/src/lxc/attach.c
> @@ -30,6 +30,7 @@
> #include <fcntl.h>
> #include <sys/param.h>
> #include <sys/prctl.h>
> +#include <sys/mount.h>
> #include <linux/unistd.h>
>
> #if !HAVE_DECL_PR_CAPBSET_DROP
> @@ -188,6 +189,49 @@ int lxc_attach_to_ns(pid_t pid, int which)
> return 0;
> }
>
> +int lxc_attach_remount_sys_proc()
> +{
> + int ret;
> +
> + ret = unshare(CLONE_NEWNS);
> + if (ret < 0) {
> + SYSERROR("failed to unshare mount namespace");
> + return -1;
> + }
> +
> + /* assume /proc is always mounted, so remount it */
> + ret = umount2("/proc", MNT_DETACH);
> + if (ret < 0) {
> + SYSERROR("failed to unmount /proc");
> + return -1;
> + }
> +
> + ret = mount("none", "/proc", "proc", 0, NULL);
> + if (ret < 0) {
> + SYSERROR("failed to remount /proc");
> + return -1;
> + }
> +
> + /* try to umount /sys - if it's not a mount point,
> + * we'll get EINVAL, then we ignore it because it
> + * may not have been mounted in the first place
> + */
> + ret = umount2("/sys", MNT_DETACH);
> + if (ret < 0 && errno != EINVAL) {
> + SYSERROR("failed to unmount /sys");
> + return -1;
> + } else if (ret == 0) {
> + /* remount it */
> + ret = mount("none", "/sys", "sysfs", 0, NULL);
> + if (ret < 0) {
> + SYSERROR("failed to remount /sys");
> + return -1;
> + }
> + }
> +
> + return 0;
> +}
> +
> int lxc_attach_drop_privs(struct lxc_proc_context_info *ctx)
> {
> int last_cap = lxc_caps_last_cap();
> diff --git a/src/lxc/attach.h b/src/lxc/attach.h
> index d96fdae..aab47e3 100644
> --- a/src/lxc/attach.h
> +++ b/src/lxc/attach.h
> @@ -34,6 +34,7 @@ struct lxc_proc_context_info {
> extern struct lxc_proc_context_info *lxc_proc_get_context_info(pid_t pid);
>
> extern int lxc_attach_to_ns(pid_t other_pid, int which);
> +extern int lxc_attach_remount_sys_proc();
> extern int lxc_attach_drop_privs(struct lxc_proc_context_info *ctx);
>
> #endif
> diff --git a/src/lxc/lxc_attach.c b/src/lxc/lxc_attach.c
> index 4f22752..e292bc4 100644
> --- a/src/lxc/lxc_attach.c
> +++ b/src/lxc/lxc_attach.c
> @@ -48,12 +48,14 @@ static const struct option my_longopts[] = {
> {"elevated-privileges", no_argument, 0, 'e'},
> {"arch", required_argument, 0, 'a'},
> {"namespaces", required_argument, 0, 's'},
> + {"remount-sys-proc", no_argument, 0, 'R'},
> LXC_COMMON_OPTIONS
> };
>
> static int elevated_privileges = 0;
> static signed long new_personality = -1;
> static int namespace_flags = -1;
> +static int remount_sys_proc = 0;
>
> static int my_parser(struct lxc_arguments* args, int c, char* arg)
> {
> @@ -61,6 +63,7 @@ static int my_parser(struct lxc_arguments* args, int c, char* arg)
>
> switch (c) {
> case 'e': elevated_privileges = 1; break;
> + case 'R': remount_sys_proc = 1; break;
> case 'a':
> new_personality = lxc_config_parse_arch(arg);
> if (new_personality < 0) {
> @@ -102,7 +105,12 @@ Options :\n\
> but just to the following OR'd list of flags:\n\
> MOUNT, PID, UTSNAME, IPC, USER or NETWORK\n\
> WARNING: Using -s implies -e, it may therefore\n\
> - leak privileges into the container. Use with care.\n",
> + leak privileges into the container. Use with care.\n\
> + -R, --remount-sys-proc\n\
> + Remount /sys and /proc if not attaching to the\n\
> + mount namespace when using -s in order to properly\n\
> + reflect the correct namespace context. See the\n\
> + lxc-attach(1) manual page for details.\n",
> .options = my_longopts,
> .parser = my_parser,
> .checker = NULL,
> @@ -253,6 +261,18 @@ int main(int argc, char *argv[])
> lxc_sync_fini_parent(handler);
> lxc_cgroup_dispose_attach(cgroup_data);
>
> + /* A description of the purpose of this functionality is
> + * provided in the lxc-attach(1) manual page. We have to
> + * remount here and not in the parent process, otherwise
> + * /proc may not properly reflect the new pid namespace.
> + */
> + if (!(namespace_flags & CLONE_NEWNS) && remount_sys_proc) {
> + ret = lxc_attach_remount_sys_proc();
> + if (ret < 0) {
> + return -1;
> + }
> + }
> +
> if (new_personality < 0)
> new_personality = init_ctx->personality;
>
> --
> 1.7.2.5
>
>
> ------------------------------------------------------------------------------
> Live Security Virtual Conference
> Exclusive live event will cover all the ways today's security and
> threat landscape has changed and how IT managers can respond. Discussions
> will include endpoint security, mobile security and the latest in malware
> threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
> _______________________________________________
> Lxc-devel mailing list
> Lxc-devel at lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/lxc-devel
More information about the lxc-devel
mailing list