Messages in this thread | | | From | Witold Krecicki <> | Subject | [PATCH 0/6] cgroup: add isolation_root flag, poor man's namespaces for cgroups | Date | Fri, 30 Sep 2011 13:36:19 +0200 |
| |
This patchset adds namespace-like feature to the existing cgroup system. When used with a container system (eg. lxc) it allows containers to have its own cgroup hierarchy, enabling use of 'systemd' (using cgroups) inside a container.
The behaviour is not changed at all when isolation_root flag is not set, no additional overhead is added.
I'm really not sure if the 'mount' part (patch 5) is done correctly, please review carefully.
Witold Krecicki (6): cgroup: add cgroup.isolation_root flag entry to the cgroup filesystem cgroup: make 'cgroup_is_descendant' function take cgroup as a 'descendant of' argument cgroup: add 'root' parameter to cgroup_path function cgroup: disallow task from leaving cgroup isolated root cgroup: make cgroup filesystem mounts performed by task inside isolation root see its isolation root as top cgroup cgroup: documentation of isolation_root cgroup flag
Documentation/cgroups/cgroups.txt | 20 +++++++- block/blk-cgroup.c | 2 +- include/linux/cgroup.h | 12 +++- kernel/cgroup.c | 102 +++++++++++++++++++++++++++++++----- kernel/cpuset.c | 2 +- kernel/sched_debug.c | 2 +- mm/memcontrol.c | 6 +- 7 files changed, 120 insertions(+), 26 deletions(-)
-- 1.7.4.1
| |