| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This patch handles the validation of the newer configuration files.
Some of the rules to be followed
1. We cannot have more controllers in the namespcae section than
already mounted.
2. If more than one controller are mounted at the same point, then
they will have the same namespace. In case it is not explicitly
mentioned, the subsystems at the mount point will be set to the
same namespace. This does not mean that controllers mounted at different
points need to have the same namespace.
Changes from v4:
1. Changed a variable name from mount to mount_path
2. Added more comments
Changes from v3:
1. Removed most of the strdups
2. Fixed return values for errors
Changes from v2:
1. mount and namespace keyword cannot come in the same file.
Changes from v1:
1. Fix a bug where if a namespace was not defined, we were not exiting
2. Comment the validate namespace function
3. Make some of the variables more descriptive
4. Make namespace thread specific
Signed-off-by: Dhaval Giani <dhaval@linux.vnet.ibm.com>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This patch introduces a new keywork NAMESPACE which allow cgroups
to be created starting from a certain point as mentioned by the
administrator. This is provided via the configuration file.
Newer APIs will use this feature.
Changes from v1:
1. Attempted to fix the indentation
2. Change a variable name
Signed-off-by: Dhaval Giani <dhaval@linux.vnet.ibm.com>
Acked-by: Balbir Singh <balbir@linux.vnet.ibm.com>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Changelog since v2:
o Avoid breaking ABI by changing the ECGCONFIGPARSEFAIL value.
Changelog since v1:
o Change a parser error value of /etc/cgconfig.conf to ECGCONFIGPARSEFAIL.
The cgconfig service succeeds even if there are something wrong
in /etc/cgconfig.conf. For example, it succeeds even if specifying
not-existing user (nouser) as a perm.task.uid like the following:
# cat /etc/cgconfig.conf
mount {
cpuset = /mnt/cgroups/cpuset;
memory = /mnt/cgroups/memory;
}
group root {
perm {
task {
uid = nouser;
gid = root;
}
admin {
uid = root;
gid = root;
}
}
cpuset {
cpuset.cpus = 0;
cpuset.mems = 0;
}
memory {
memory.use_hierarchy = 1;
memory.limit_in_bytes = 1000000000000;
}
}
#
# service cgconfig start
Starting cgconfig service: parsing failed at line number 10
[ OK ]
# echo $?
0
#
That is not a good behavior and the cgconfig service should fail
if there are something wrong in /etc/cgconfig.conf.
This patch fixes the behavior. In the above case, the behavior is
changed like the following by this patch:
# service cgconfig start
Starting cgconfig service: parsing failed at line number 10
Loading configuration file /etc/cgconfig.conf failed
Cgroup parsing failed
Failed to parse /etc/cgconfig.conf [FAILED]
#
Signed-off-by: Ken'ichi Ohmichi <oomichi@mxs.nes.nec.co.jp>
Signed-off-by: Dhaval Giani <dhaval@linux.vnet.ibm.com>
|
| |
|
|
|
|
|
|
|
|
|
|
| |
Let's rework cgroup_config_unload_controller to use the brand new recursive
removal of groups. It saves *lot of* fopen() and fclose() calls:
On my (semi-broken) Fedora 11 with 196 running processes and 3 cgroup
hierarchies, the old cgclear does 668 open("*/tasks"). With the
optimized cgclear, I can see just 22 open("*/tasks") calls.
Signed-off-by: Jan Safranek <jsafrane@redhat.com>
Signed-off-by: Dhaval Giani <dhaval@linux.vnet.ibm.com>
|
| |
|
|
|
|
|
| |
Jan pointed out that there was a goto missing in the error
handling paths. Fix that bug.
Signed-off-by: Dhaval Giani <dhaval@linux.vnet.ibm.com>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This API will unload the cgroups created in the cgroupfs and
unmount and delete the filesystem mount point. The action is
equivalent to what is done currently in service cgconfig stop.
The reason for this API is to make sure we don't end up with a
asymmetric library API subset. Today an application program can
programatically through cgroup_config_load_config() load a
configuration file, but has no means to cleanup (including all
temporarily created groups).
changes from v3
1. Address Jan's comments from http://article.gmane.org/gmane.comp.lib.libcg.devel/1105
changes from v2
1. Fix a leak as noted by Bharata
2. Address Balbir's review comments at
http://article.gmane.org/gmane.comp.lib.libcg.devel/1080
changes from v1
1. Change the name of the function to cgroup_unload_cgroups
2. Change the name of the executatble to cgclear
3. Split out the funtions
Signed-off-by: Dhaval Giani <dhaval@linux.vnet.ibm.com>
Acked-by: Balbir Singh <balbir@linux.vnet.ibm.com>
|
| |
|
|
|
|
|
|
|
|
|
| |
https://bugzilla.redhat.com/show_bug.cgi?id=502687 mentioned that the directory
was not getting created when a cgconfig start was being run.
This is was because we failed the mkdir. The mkdir for directories at depth
was not succeeding.
Signed-off-by: Dhaval Giani <dhaval@linux.vnet.ibm.com>
Acked-by: Balbir Singh <balbir@linux.vnet.ibm.com>
|
| |
|
|
|
|
|
|
|
| |
I reviewed the libcgroup code in the viewpoint of memory-leak and
found some lacks of fclose(). This patch adds some fclose() calls
for error handling.
Signed-off-by: Ken'ichi Ohmichi <oomichi@mxs.nes.nec.co.jp>
Signed-off-by: Dhaval Giani <dhaval@linux.vnet.ibm.com>
|
|
|
Signed-off-by: Jan Safranek <jsafrane@redhat.com>
|
