Skip to contentSkip to conversation

Containers (V9): Generic Process Containers

PATCH In review [email protected] 5 participants 36 messages Apr 27, 2007 – May 11, 2007

Subsystems (8)
lore
Share & tools
lore
#0 lore raw

cover Containers (V9): Generic Process Containers

--

This is an update to my multi-hierarchy multi-subsystem generic process containers patch. Changes since V8 (April 6th) include:

- The patchset has been rebased over 2.6.21-rc7-mm1

- The patchset has been restructured based on feedback; more functionality is now split out into separate patches where practical.

- The container_group structure has been renamed css_group since this is more descriptive of its true function

- Added a simplified file registration interface, and a simple interface for the common operation of returning a single number to userspace from a container control file

- Added a simple "debug" subsystem that is both an example of how to use the container system and a useful debugging tool for checking reference counts, etc.

Still TODO:

- decide whether "Containers" is an acceptable name for the system given its usage by some other development groups, or whether something else (ProcessSets? ResourceGroups? TaskGroups?) would be better

- decide whether merging css_group and nsproxy is desirable

- add a hash-table based lookup for css_group objects.

- use seq_file properly in container tasks files to avoid having to allocate a big array for all the container's task pointers.

- add back support for the "release agent" functionality

- lots more testing

- define standards for container file names

Generic Process Containers
--------------------------

There have recently been various proposals floating around for resource management/accounting and other task grouping subsystems in the kernel, including ResGroups, User BeanCounters, NSProxy containers, and others. These all need the basic abstraction of being able to group together multiple processes in an aggregate, in order to track/limit the resources permitted to those processes, or control other behaviour of the processes, and all implement this grouping in different ways.

Already existing in the kernel is the cpuset subsystem; this has a process grouping mechanism that is mature, tested, and well documented (particularly with regards to synchronization rules).

This patchset extracts the process grouping code from cpusets into a generic container system, and makes the cpusets code a client of the container system, along with a couple of simple example subsystems.

The patch set is structured as follows:

1) Basic container framework - filesystem and tracking structures

2) Simple CPU Accounting example subsystem

3) Support for the "tasks" control file

4) Hooks for fork() and exit()

5) Support for the container_clone() operation

6) Add /proc reporting interface

7) Make cpusets a container subsystem

8) Share container subsystem pointer arrays between tasks with the same assignments

9) Simple container debugging subsystem

The intention is that the various resource management and virtualization efforts can also become container clients, with the result that:

- the userspace APIs are (somewhat) normalised

- it's easier to test out e.g. the ResGroups CPU controller in conjunction with the BeanCounters memory controller, or use either of
them as the resource-control portion of a virtual server system.

- the additional kernel footprint of any of the competing resource management systems is substantially reduced, since it doesn't need to provide process grouping/containment, hence improving their chances of getting into the kernel

#1 lore raw

PATCH 1/9 Containers (V9): Basic container framework

This patch adds the main containers framework - the container filesystem, and the basic structures for tracking membership and associating subsystem state objects to tasks.

---
 Documentation/containers.txt     |  524 +++++++++++++++++
 include/linux/container.h        |  198 ++++++
 include/linux/container_subsys.h |   10 
 include/linux/sched.h            |   34 +
 init/Kconfig                     |    3 
 init/main.c                      |    3 
 kernel/Makefile                  |    1 
 kernel/container.c               | 1151 +++++++++++++++++++++++++++++++++++++++
 8 files changed, 1923 insertions(+), 1 deletion(-)
Documentation/containers.txt new file +524 -0
Index: container-2.6.21-rc7-mm1/Documentation/containers.txt
===================================================================
@@ -0,0 +1,524 @@
+				CONTAINERS
+				-------
+
+Written by Paul Menage <[email protected]> based on Documentation/cpusets.txt
+
+Original copyright statements from cpusets.txt:
+Portions Copyright (C) 2004 BULL SA.
+Portions Copyright (c) 2004-2006 Silicon Graphics, Inc.
+Modified by Paul Jackson <[email protected]>
+Modified by Christoph Lameter <[email protected]>
+
+CONTENTS:
+=========
+
+1. Containers
+  1.1 What are containers ?
+  1.2 Why are containers needed ?
+  1.3 How are containers implemented ?
+  1.4 What does notify_on_release do ?
+  1.5 How do I use containers ?
+2. Usage Examples and Syntax
+  2.1 Basic Usage
+  2.2 Attaching processes
+3. Kernel API
+  3.1 Overview
+  3.2 Synchronization
+  3.3 Subsystem API
+4. Questions
+
+1. Containers
+==========
+
+1.1 What are containers ?
+----------------------
+
+Containers provide a mechanism for aggregating/partitioning sets of
+tasks, and all their future children, into hierarchical groups with
+specialized behaviour.
+
+Definitions:
+
+A *container* associates a set of tasks with a set of parameters for one
+or more subsystems.
+
+A *subsystem* is a module that makes use of the task grouping
+facilities provided by containers to treat groups of tasks in
+particular ways. A subsystem is typically a "resource controller" that
+schedules a resource or applies per-container limits, but it may be
+anything that wants to act on a group of processes, e.g. a
+virtualization subsystem.
+
+A *hierarchy* is a set of containers arranged in a tree, such that
+every task in the system is in exactly one of the containers in the
+hierarchy, and a set of subsystems; each subsystem has system-specific
+state attached to each container in the hierarchy.  Each hierarchy has
+an instance of the container virtual filesystem associated with it.
+
+At any one time there may be multiple active hierachies of task
+containers. Each hierarchy is a partition of all tasks in the system.
+
+User level code may create and destroy containers by name in an
+instance of the container virtual file system, specify and query to
+which container a task is assigned, and list the task pids assigned to
+a container. Those creations and assignments only affect the hierarchy
+associated with that instance of the container file system.
+
+On their own, the only use for containers is for simple job
+tracking. The intention is that other subsystems hook into the generic
+container support to provide new attributes for containers, such as
+accounting/limiting the resources which processes in a container can
+access. For example, cpusets (see Documentation/cpusets.txt) allows
+you to associate a set of CPUs and a set of memory nodes with the
+tasks in each container.
+
+1.2 Why are containers needed ?
+----------------------------
+
+There are multiple efforts to provide process aggregations in the
+Linux kernel, mainly for resource tracking purposes. Such efforts
+include cpusets, CKRM/ResGroups, UserBeanCounters, and virtual server
+namespaces. These all require the basic notion of a
+grouping/partitioning of processes, with newly forked processes ending
+in the same group (container) as their parent process.
+
+The kernel container patch provides the minimum essential kernel
+mechanisms required to efficiently implement such groups. It has
+minimal impact on the system fast paths, and provides hooks for
+specific subsystems such as cpusets to provide additional behaviour as
+desired.
+
+Multiple hierarchy support is provided to allow for situations where
+the division of tasks into containers is distinctly different for
+different subsystems - having parallel hierarchies allows each
+hierarchy to be a natural division of tasks, without having to handle
+complex combinations of tasks that would be present if several
+unrelated subsystems needed to be forced into the same tree of
+containers.
+
+At one extreme, each resource controller or subsystem could be in a
+separate hierarchy; at the other extreme, all subsystems
+would be attached to the same hierarchy.
+
+As an example of a scenario (originally proposed by [email protected])
+that can benefit from multiple hierarchies, consider a large
+university server with various users - students, professors, system
+tasks etc. The resource planning for this server could be along the
+following lines:
+
+       CPU :           Top cpuset
+                       /       \
+               CPUSet1         CPUSet2
+                  |              |
+               (Profs)         (Students)
+
+               In addition (system tasks) are attached to topcpuset (so
+               that they can run anywhere) with a limit of 20%
+
+       Memory : Professors (50%), students (30%), system (20%)
+
+       Disk : Prof (50%), students (30%), system (20%)
+
+       Network : WWW browsing (20%), Network File System (60%), others (20%)
+                               / \
+                       Prof (15%) students (5%)
+
+Browsers like firefox/lynx go into the WWW network class, while (k)nfsd go
+into NFS network class.
+
+At the same time firefox/lynx will share an appropriate CPU/Memory class
+depending on who launched it (prof/student).
+
+With the ability to classify tasks differently for different resources
+(by putting those resource subsystems in different hierarchies) then
+the admin can easily set up a script which receives exec notifications
+and depending on who is launching the browser he can
+
+       # echo browser_pid > /mnt/<restype>/<userclass>/tasks
+
+With only a single hierarchy, he now would potentially have to create
+a separate container for every browser launched and associate it with
+approp network and other resource class.  This may lead to
+proliferation of such containers.
+
+Also lets say that the administrator would like to give enhanced network
+access temporarily to a student's browser (since it is night and the user
+wants to do online gaming :)  OR give one of the students simulation
+apps enhanced CPU power,
+
+With ability to write pids directly to resource classes, its just a
+matter of :
+
+       # echo pid > /mnt/network/<new_class>/tasks
+       (after some time)
+       # echo pid > /mnt/network/<orig_class>/tasks
+
+Without this ability, he would have to split the container into
+multiple separate ones and then associate the new containers with the
+new resource classes.
+
+
+
+1.3 How are containers implemented ?
+---------------------------------
+
+Containers extends the kernel as follows:
+
+ - Each task in the system has a reference-counted pointer to a
+   css_group.
+
+ - A css_group contains a set of reference-counted pointers to
+   container_subsys_state objects, one for each container subsystem
+   registered in the system. There is no direct link from a task to
+   the container of which it's a member in each hierarchy, but this
+   can be determined by following pointers through the
+   container_subsys_state objects. This is because accessing the
+   subsystem state is something that's expected to happen frequently
+   and in performance-critical code, whereas operations that require a
+   task's actual container assignments (in particular, moving between
+   containers) are less common.
+
+ - A container hierarchy filesystem can be mounted  for browsing and
+   manipulation from user space.
+
+ - You can list all the tasks (by pid) attached to any container.
+
+The implementation of containers requires a few, simple hooks
+into the rest of the kernel, none in performance critical paths:
+
+ - in init/main.c, to initialize the root containers and initial
+   css_group at system boot.
+
+ - in fork and exit, to attach and detach a task from its css_group.
+
+In addition a new file system, of type "container" may be mounted, to
+enable browsing and modifying the containers presently known to the
+kernel.  When mounting a container hierarchy, you may specify a
+comma-separated list of subsystems to mount as the filesystem mount
+options.  By default, mounting the container filesystem attempts to
+mount a hierarchy containing all registered subsystems.
+
+If an active hierarchy with exactly the same set of subsystems already
+exists, it will be reused for the new mount. If no existing hierarchy
+matches, and any of the requested subsystems are in use in an existing
+hierarchy, the mount will fail with -EBUSY. Otherwise, a new hierarchy
+is activated, associated with the requested subsystems.
+
+It's not currently possible to bind a new subsystem to an active
+container hierarchy, or to unbind a subsystem from an active container
+hierarchy. This may be possible in future, but is fraught with nasty
+error-recovery issues.
+
+When a container filesystem is unmounted, if there are any
+subcontainers created below the top-level container, that hierarchy
+will remain active even though unmounted; if there are no
+subcontainers then the hierarchy will be deactivated.
+
+No new system calls are added for containers - all support for
+querying and modifying containers is via this container file system.
+
+Each task under /proc has an added file named 'container' displaying,
+for each active hierarchy, the subsystem names and the container name
+as the path relative to the root of the container file system.
+
+Each container is represented by a directory in the container file system
+containing the following files describing that container:
+
+ - tasks: list of tasks (by pid) attached to that container
+ - notify_on_release flag: run /sbin/container_release_agent on exit?
+
+Other subsystems such as cpusets may add additional files in each
+container dir
+
+New containers are created using the mkdir system call or shell
+command.  The properties of a container, such as its flags, are
+modified by writing to the appropriate file in that containers
+directory, as listed above.
+
+The named hierarchical structure of nested containers allows partitioning
+a large system into nested, dynamically changeable, "soft-partitions".
+
+The attachment of each task, automatically inherited at fork by any
+children of that task, to a container allows organizing the work load
+on a system into related sets of tasks.  A task may be re-attached to
+any other container, if allowed by the permissions on the necessary
+container file system directories.
+
+When a task is moved from one container to another, it gets a new
+css_group pointer - if there's an already existing css_group with the
+desired collection of containers then that group is reused, else a new
+css_group is allocated. Note that the current implementation uses a
+linear search to locate an appropriate existing css_group, so isn't
+very efficient. A future version will use a hash table for better
+performance.
+
+The use of a Linux virtual file system (vfs) to represent the
+container hierarchy provides for a familiar permission and name space
+for containers, with a minimum of additional kernel code.
+
+1.4 What does notify_on_release do ?
+------------------------------------
+
+*** notify_on_release is disabled in the current patch set. It may be
+*** reactivated in a future patch in a less-intrusive manner
+
+If the notify_on_release flag is enabled (1) in a container, then
+whenever the last task in the container leaves (exits or attaches to
+some other container) and the last child container of that container
+is removed, then the kernel runs the command specified by the contents
+of the "release_agent" file in that hierarchy's root directory,
+supplying the pathname (relative to the mount point of the container
+file system) of the abandoned container.  This enables automatic
+removal of abandoned containers.  The default value of
+notify_on_release in the root container at system boot is disabled
+(0).  The default value of other containers at creation is the current
+value of their parents notify_on_release setting. The default value of
+a container hierarchy's release_agent path is empty.
+
+1.5 How do I use containers ?
+--------------------------
+
+To start a new job that is to be contained within a container, using
+the "cpuset" container subsystem, the steps are something like:
+
+ 1) mkdir /dev/container
+ 2) mount -t container -ocpuset cpuset /dev/container
+ 3) Create the new container by doing mkdir's and write's (or echo's) in
+    the /dev/container virtual file system.
+ 4) Start a task that will be the "founding father" of the new job.
+ 5) Attach that task to the new container by writing its pid to the
+    /dev/container tasks file for that container.
+ 6) fork, exec or clone the job tasks from this founding father task.
+
+For example, the following sequence of commands will setup a container
+named "Charlie", containing just CPUs 2 and 3, and Memory Node 1,
+and then start a subshell 'sh' in that container:
+
+  mount -t container cpuset -ocpuset /dev/container
+  cd /dev/container
+  mkdir Charlie
+  cd Charlie
+  /bin/echo $$ > tasks
+  sh
+  # The subshell 'sh' is now running in container Charlie
+  # The next line should display '/Charlie'
+  cat /proc/self/container
+
+2. Usage Examples and Syntax
+============================
+
+2.1 Basic Usage
+---------------
+
+Creating, modifying, using the containers can be done through the container
+virtual filesystem.
+
+To mount a container hierarchy will all available subsystems, type:
+# mount -t container xxx /dev/container
+
+The "xxx" is not interpreted by the container code, but will appear in
+/proc/mounts so may be any useful identifying string that you like.
+
+To mount a container hierarchy with just the cpuset and numtasks
+subsystems, type:
+# mount -t container -o cpuset,numtasks hier1 /dev/container
+
+To change the set of subsystems bound to a mounted hierarchy, just
+remount with different options:
+
+# mount -o remount,cpuset,ns  /dev/container
+
+Note that changing the set of subsystems is currently only supported
+when the hierarchy consists of a single (root) container. Supporting
+the ability to arbitrarily bind/unbind subsystems from an existing
+container hierarchy is intended to be implemented in the future.
+
+Then under /dev/container you can find a tree that corresponds to the
+tree of the containers in the system. For instance, /dev/container
+is the container that holds the whole system.
+
+If you want to create a new container under /dev/container:
+# cd /dev/container
+# mkdir my_container
+
+Now you want to do something with this container.
+# cd my_container
+
+In this directory you can find several files:
+# ls
+notify_on_release release_agent tasks
+(plus whatever files are added by the attached subsystems)
+
+Now attach your shell to this container:
+# /bin/echo $$ > tasks
+
+You can also create containers inside your container by using mkdir in this
+directory.
+# mkdir my_sub_cs
+
+To remove a container, just use rmdir:
+# rmdir my_sub_cs
+
+This will fail if the container is in use (has containers inside, or
+has processes attached, or is held alive by other subsystem-specific
+reference).
+
+2.2 Attaching processes
+-----------------------
+
+# /bin/echo PID > tasks
+
+Note that it is PID, not PIDs. You can only attach ONE task at a time.
+If you have several tasks to attach, you have to do it one after another:
+
+# /bin/echo PID1 > tasks
+# /bin/echo PID2 > tasks
+	...
+# /bin/echo PIDn > tasks
+
+3. Kernel API
+=============
+
+3.1 Overview
+------------
+
+Each kernel subsystem that wants to hook into the generic container
+system needs to create a container_subsys object. This contains
+various methods, which are callbacks from the container system, along
+with a subsystem id which will be assigned by the container system.
+
+Other fields in the container_subsys object include:
+
+- subsys_id: a unique array index for the subsystem, indicating which
+  entry in container->subsys[] this subsystem should be
+  managing. Initialized by container_register_subsys(); prior to this
+  it should be initialized to -1
+
+- hierarchy: an index indicating which hierarchy, if any, this
+  subsystem is currently attached to. If this is -1, then the
+  subsystem is not attached to any hierarchy, and all tasks should be
+  considered to be members of the subsystem's top_container. It should
+  be initialized to -1.
+
+- name: should be initialized to a unique subsystem name prior to
+  calling container_register_subsystem. Should be no longer than
+  MAX_CONTAINER_TYPE_NAMELEN
+
+Each container object created by the system has an array of pointers,
+indexed by subsystem id; this pointer is entirely managed by the
+subsystem; the generic container code will never touch this pointer.
+
+3.2 Synchronization
+-------------------
+
+There is a global mutex, container_mutex, used by the container
+system. This should be taken by anything that wants to modify a
+container. It may also be taken to prevent containers from being
+modified, but more specific locks may be more appropriate in that
+situation.
+
+See kernel/container.c for more details.
+
+Subsystems can take/release the container_mutex via the functions
+container_lock()/container_unlock(), and can
+take/release the callback_mutex via the functions
+container_lock()/container_unlock().
+
+Accessing a task's container pointer may be done in the following ways:
+- while holding container_mutex
+- while holding the task's alloc_lock (via task_lock())
+- inside an rcu_read_lock() section via rcu_dereference()
+
+3.3 Subsystem API
+--------------------------
+
+Each subsystem should:
+
+- add an entry in linux/container_subsys.h
+- define a container_subsys object called <name>_subsys
+
+Each subsystem may export the following methods. The only mandatory
+methods are create/destroy. Any others that are null are presumed to
+be successful no-ops.
+
+int create(struct container *cont)
+LL=container_mutex
+
+Called to create a subsystem state object for a container. The
+subsystem should set its subsystem pointer for the passed container,
+returning 0 on success or a negative error code. On success, the
+subsystem pointer should point to a structure of type
+container_subsys_state (typically embedded in a larger
+subsystem-specific object), which will be initialized by the container
+system. Note that this will be called at initialization to create the
+root subsystem state for this subsystem; this case can be identified
+by the passed container object having a NULL parent (since it's the
+root of the hierarchy) and may be an appropriate place for
+initialization code.
+
+void destroy(struct container *cont)
+LL=container_mutex
+
+The container system is about to destroy the passed container; the
+subsystem should do any necessary cleanup
+
+int can_attach(struct container_subsys *ss, struct container *cont,
+	       struct task_struct *task)
+LL=container_mutex
+
+Called prior to moving a task into a container; if the subsystem
+returns an error, this will abort the attach operation.  If a NULL
+task is passed, then a successful result indicates that *any*
+unspecified task can be moved into the container. Note that this isn't
+called on a fork. If this method returns 0 (success) then this should
+remain valid while the caller holds container_mutex.
+
+void attach(struct container_subsys *ss, struct container *cont,
+	    struct container *old_cont, struct task_struct *task)
+LL=container_mutex
+
+
+Called after the task has been attached to the container, to allow any
+post-attachment activity that requires memory allocations or blocking.
+
+void fork(struct container_subsy *ss, struct task_struct *task)
+LL=callback_mutex, maybe read_lock(tasklist_lock)
+
+Called when a task is forked into a container. Also called during
+registration for all existing tasks.
+
+void exit(struct container_subsys *ss, struct task_struct *task)
+LL=callback_mutex
+
+Called during task exit
+
+int populate(struct container_subsys *ss, struct container *cont)
+LL=none
+
+Called after creation of a container to allow a subsystem to populate
+the container directory with file entries.  The subsystem should make
+calls to container_add_file() with objects of type cftype (see
+include/linux/container.h for details).  Note that although this
+method can return an error code, the error code is currently not
+always handled well.
+
+void bind(struct container_subsys *ss, struct container *root)
+LL=callback_mutex
+
+Called when a container subsystem is rebound to a different hierarchy
+and root container. Currently this will only involve movement between
+the default hierarchy (which never has sub-containers) and a hierarchy
+that is being created/destroyed (and hence has no sub-containers).
+
+4. Questions
+============
+
+Q: what's up with this '/bin/echo' ?
+A: bash's builtin 'echo' command does not check calls to write() against
+   errors. If you use it in the container file system, you won't be
+   able to tell whether a command succeeded or failed.
+
+Q: When I attach processes, only the first of the line gets really attached !
+A: We can only return one error code per call to write(). So you should also
+   put only ONE pid.
+
include/linux/container.h new file +198 -0

198 lines not on this page · raw message

include/linux/container_subsys.h new file +10 -0
Index: container-2.6.21-rc7-mm1/include/linux/container_subsys.h
===================================================================
@@ -0,0 +1,10 @@
+/* Add subsystem definitions of the form SUBSYS(<name>) in this
+ * file. Surround each one by a line of comment markers so that
+ * patches don't collide
+ */
+
+/* */
+
+/* */
+
+/* */
include/linux/sched.h +33 -1
Index: container-2.6.21-rc7-mm1/include/linux/sched.h
===================================================================
@@ -820,6 +820,34 @@ struct uts_namespace;
 
 struct prio_array;
 
+#ifdef CONFIG_CONTAINERS
+
+#define SUBSYS(_x) _x ## _subsys_id,
+enum container_subsys_id {
+#include <linux/container_subsys.h>
+	CONTAINER_SUBSYS_COUNT
+};
+#undef SUBSYS
+
+/* A css_group is a structure holding pointers to a set of
+ * container_subsys_state objects.
+ */
+
+struct css_group {
+
+	/* Set of subsystem states, one for each subsystem. NULL for
+	 * subsystems that aren't part of this hierarchy. These
+	 * pointers reduce the number of dereferences required to get
+	 * from a task to its state for a given container, but result
+	 * in increased space usage if tasks are in wildly different
+	 * groupings across different hierarchies. This array is
+	 * immutable after creation */
+	struct container_subsys_state *subsys[CONTAINER_SUBSYS_COUNT];
+
+};
+
+#endif /* CONFIG_CONTAINERS */
+
 struct task_struct {
 	volatile long state;	/* -1 unrunnable, 0 runnable, >0 stopped */
 	struct thread_info *thread_info;
@@ -1072,6 +1100,9 @@ struct task_struct {
 	int cpuset_mems_generation;
 	int cpuset_mem_spread_rotor;
 #endif
+#ifdef CONFIG_CONTAINERS
+	struct css_group containers;
+#endif
 	struct robust_list_head __user *robust_list;
 #ifdef CONFIG_COMPAT
 	struct compat_robust_list_head __user *compat_robust_list;
@@ -1509,7 +1540,8 @@ static inline int thread_group_empty(str
 /*
  * Protects ->fs, ->files, ->mm, ->group_info, ->comm, keyring
  * subscriptions and synchronises with wait4().  Also used in procfs.  Also
- * pins the final release of task.io_context.  Also protects ->cpuset.
+ * pins the final release of task.io_context.  Also protects ->cpuset and
+ * ->container.subsys[].
  *
  * Nests both inside and outside of read_lock(&tasklist_lock).
  * It must not be nested with write_lock_irq(&tasklist_lock),
init/Kconfig +3 -0
Index: container-2.6.21-rc7-mm1/init/Kconfig
===================================================================
@@ -288,6 +288,9 @@ config IKCONFIG_PROC
 	  This option enables access to the kernel configuration file
 	  through /proc/config.gz.
 
+config CONTAINERS
+	bool
+
 config CPUSETS
 	bool "Cpuset support"
 	depends on SMP
init/main.c +3 -0
Index: container-2.6.21-rc7-mm1/init/main.c
===================================================================
@@ -39,6 +39,7 @@
 #include <linux/writeback.h>
 #include <linux/cpu.h>
 #include <linux/cpuset.h>
+#include <linux/container.h>
 #include <linux/efi.h>
 #include <linux/tick.h>
 #include <linux/interrupt.h>
@@ -499,6 +500,7 @@ asmlinkage void __init start_kernel(void
 	char * command_line;
 	extern struct kernel_param __start___param[], __stop___param[];
 
+	container_init_early();
 	smp_setup_processor_id();
 
 	/*
@@ -624,6 +626,7 @@ asmlinkage void __init start_kernel(void
 #ifdef CONFIG_PROC_FS
 	proc_root_init();
 #endif
+	container_init();
 	cpuset_init();
 	taskstats_init_early();
 	delayacct_init();
kernel/Makefile +1 -0
Index: container-2.6.21-rc7-mm1/kernel/Makefile
===================================================================
@@ -36,6 +36,7 @@ obj-$(CONFIG_PM) += power/
 obj-$(CONFIG_BSD_PROCESS_ACCT) += acct.o
 obj-$(CONFIG_KEXEC) += kexec.o
 obj-$(CONFIG_COMPAT) += compat.o
+obj-$(CONFIG_CONTAINERS) += container.o
 obj-$(CONFIG_CPUSETS) += cpuset.o
 obj-$(CONFIG_IKCONFIG) += configs.o
 obj-$(CONFIG_STOP_MACHINE) += stop_machine.o
kernel/container.c new file +1151 -0

1,151 lines not on this page · raw message

--

#2 lore raw

Re: [ckrm-tech] [PATCH 1/9] Containers (V9): Basic container framework

+static ssize_t container_read_uint(struct container *cont, struct cftype *cft,

+				   struct file *file,
+				   char __user *buf, size_t nbytes,
+				   loff_t *ppos)

+{
+ char tmp[64];
+ u64 val = cft->read_uint(cont, cft);
+ int len = sprintf(tmp, "%llu", val);

This sprintf is giving me the following compiler warnings:

kernel/container.c: In function 'container_read_uint':
kernel/container.c:1025: warning: format '%llu' expects type 'long long unsigned int', but argument 3 has type 'u64' kernel/container.c:1025: warning: format '%llu' expects type 'long long unsigned int', but argument 3 has type 'u64'

Signature · 3 lines
                  I won't rest till it's the best ...
                  Programmer, Linux Scalability
                  Paul Jackson <[email protected]> 1.925.600.0401
#3 lore raw

Re: [PATCH 1/9] Containers (V9): Basic container framework

In reply to [email protected]
Quoted 3 lines · [email protected] wrote:

This patch adds the main containers framework - the container filesystem, and the basic structures for tracking membership and associating subsystem state objects to tasks.

[snip]

Documentation/containers.txt line 264

Quoted 3 lines
Documentation/containers.txt new file +3 -0
+*** notify_on_release is disabled in the current patch set. It may be
+*** reactivated in a future patch in a less-intrusive manner
+

Won't this break user space tools for cpusets?

[snip]

Documentation/containers.txt line 426

Quoted 7 lines
Documentation/containers.txt new file +7 -0
+See kernel/container.c for more details.
+
+Subsystems can take/release the container_mutex via the functions
+container_lock()/container_unlock(), and can
+take/release the callback_mutex via the functions
+container_lock()/container_unlock().
+

Hmm.. looks like a documentation error. Both mutex's are obtained through container_lock/container_unlock ?

Documentation/containers.txt line 431

Quoted 5 lines
Documentation/containers.txt new file +5 -0
+Accessing a task's container pointer may be done in the following ways:
+- while holding container_mutex
+- while holding the task's alloc_lock (via task_lock())
+- inside an rcu_read_lock() section via rcu_dereference()
+

container_mutex() and task_lock() can be used for changing the pointer? Could you please explain this a bit further.

[snip]

Documentation/containers.txt line 503

Quoted 9 lines
Documentation/containers.txt new file +9 -0
+int populate(struct container_subsys *ss, struct container *cont)
+LL=none
+
+Called after creation of a container to allow a subsystem to populate
+the container directory with file entries.  The subsystem should make
+calls to container_add_file() with objects of type cftype (see
+include/linux/container.h for details).  Note that although this
+method can return an error code, the error code is currently not
+always handled well.

We needed the equivalent of container_remove_file() to be called if container_add_file() failed.

[snip]

Quoted 14 lines
include/linux/container.h new file +14 -0
+struct container {
+	unsigned long flags;		/* "unsigned long" so bitops work */
+
+	/* count users of this container. >0 means busy, but doesn't
+	 * necessarily indicate the number of tasks in the
+	 * container */
+	atomic_t count;
+
+	/*
+	 * We link our 'sibling' struct into our parent's 'children'.
+	 * Our children link their 'sibling' into our 'children'.
+	 */
+	struct list_head sibling;	/* my parent's children */
+	struct list_head children;	/* my children */

include/linux/container.h line 85

Quoted 10 lines
include/linux/container.h new file +10 -0
+
+	struct container *parent;	/* my parent */
+	struct dentry *dentry;	  	/* container fs entry */
+
+	/* Private pointers for each registered subsystem */
+	struct container_subsys_state *subsys[CONTAINER_SUBSYS_COUNT];
+
+	struct containerfs_root *root;
+	struct container *top_container;
+};

Can't we derive the top_container from containerfs_root?

Quoted 16 lines
include/linux/container.h new file +16 -0
+
+/* struct cftype:
+ *
+ * The files in the container filesystem mostly have a very simple read/write
+ * handling, some common function will take care of it. Nevertheless some cases
+ * (read tasks) are special and therefore I define this structure for every
+ * kind of file.
+ *
+ *
+ * When reading/writing to a file:
+ *	- the container to use in file->f_dentry->d_parent->d_fsdata
+ *	- the 'cftype' of the file is file->f_dentry->d_fsdata
+ */
+
+struct inode;
+#define MAX_CFTYPE_NAME 64

include/linux/container.h line 111

Quoted 10 lines
include/linux/container.h new file +10 -0
+struct cftype {
+	/* By convention, the name should begin with the name of the
+	 * subsystem, followed by a period */
+	char name[MAX_CFTYPE_NAME];
+	int private;
+	int (*open) (struct inode *inode, struct file *file);
+	ssize_t (*read) (struct container *cont, struct cftype *cft,
+			 struct file *file,
+			 char __user *buf, size_t nbytes, loff_t *ppos);
+	u64 (*read_uint) (struct container *cont, struct cftype *cft);

Is this a new callback, a specialization of the read() callback?

Quoted 6 lines
include/linux/container.h new file +6 -0
+	ssize_t (*write) (struct container *cont, struct cftype *cft,
+			  struct file *file,
+			  const char __user *buf, size_t nbytes, loff_t *ppos);
+	int (*release) (struct inode *inode, struct file *file);
+};
+

[snip]

Signature · 4 lines
	Warm Regards,
	Balbir Singh
	Linux Technology Center
	IBM, ISTL
#4 lore raw

Re: [PATCH 1/9] Containers (V9): Basic container framework

Quoted 5 lines · Balbir Singh
Quoted 3 lines · [email protected] wrote:

This patch adds the main containers framework - the container filesystem, and the basic structures for tracking membership and associating subsystem state objects to tasks.

Quoted 7 lines

[snip]

Documentation/containers.txt line 264

Quoted 3 lines
Documentation/containers.txt new file +3 -0
+*** notify_on_release is disabled in the current patch set. It may be
+*** reactivated in a future patch in a less-intrusive manner
+

Won't this break user space tools for cpusets?

Yes, so it's a must-fix before this gets anywhere near a real distribution.

Quoted 13 lines

[snip]

Documentation/containers.txt line 426

Quoted 7 lines
Documentation/containers.txt new file +7 -0
+See kernel/container.c for more details.
+
+Subsystems can take/release the container_mutex via the functions
+container_lock()/container_unlock(), and can
+take/release the callback_mutex via the functions
+container_lock()/container_unlock().
+

Hmm.. looks like a documentation error. Both mutex's are obtained through container_lock/container_unlock ?

The second half of that sentence is obsolete and should have been deleted.

Quoted 8 lines

Documentation/containers.txt line 431

Quoted 6 lines
Documentation/containers.txt new file +5 -0
+Accessing a task's container pointer may be done in the following ways:
+- while holding container_mutex
+- while holding the task's alloc_lock (via task_lock())
+- inside an rcu_read_lock() section via rcu_dereference()
+

container_mutex() and task_lock() can be used for changing the pointer?

No, these are all for read operations. (Actually, this is a bit of documentation that's bit-rotted - there's no longer a per-task "container" pointer). I'll update this.

For write operations, only the container system should be modifying those pointers (under the protection of both container_mutex and alloc_lock).

Quoted 4 lines

We needed the equivalent of container_remove_file() to be called if container_add_file() failed.

Yes, this is some incomplete behaviour that I inherited from cpusets. Needs tidying up.

Quoted 2 lines

Can't we derive the top_container from containerfs_root?

Yes, we could for the cost of an extra dereference. Not sure it's a big deal either way.

Quoted 6 lines

include/linux/container.h line 111

Quoted 4 lines
include/linux/container.h new file +4 -0
+     ssize_t (*read) (struct container *cont, struct cftype *cft,
+                      struct file *file,
+                      char __user *buf, size_t nbytes, loff_t *ppos);
+     u64 (*read_uint) (struct container *cont, struct cftype *cft);

Is this a new callback, a specialization of the read() callback?

Yes. It's to simplify the common case of reporting a number in a control file. (Not yet well documented :-( )

Paul

#5 lore raw

Re: [ckrm-tech] [PATCH 1/9] Containers (V9): Basic container framework

In reply to Balbir Singh

[[ I have bcc'd one or more batch scheduler experts on this post.
They will know who they are, and should be aware that they are
not listed in the public cc list of this message. - pj ]]

Quoted 6 lines · responding to Paul Menage's Container patch set on lkml,

Documentation/containers.txt line 264

Quoted 4 lines
Documentation/containers.txt new file +3 -0
+*** notify_on_release is disabled in the current patch set. It may be
+*** reactivated in a future patch in a less-intrusive manner
+

Won't this break user space tools for cpusets?

Yes - disabling notify_on_release would definitely break some important uses of cpusets. This feature must be reactivated somehow before I'll sign up for putting this patch set in the main line.

Actually, after I posted a few days ago in another lkml post:

	http://lkml.org/lkml/2007/4/29/66

that just the simplest cpuset command:
mount -t cpuset cpuset /dev/cpuset
mkdir /dev/cpuset/foo
echo 0 > /dev/cpuset/foo/mems

caused an immediate kernel deadlock (Srivatsa has proposed a fix), it is pretty clear that this container patch set is not getting the cpuset testing it will need for acceptance. That's partly my fault.

The batch scheduler folks, such as the variants of PBS, LSF and SGE are major user of cpusets on NUMA hardware.

This container based replacement for cpusets isn't ready for the main line until at least one of those schedulers can run through one of their test suites. I hesitate to even acknowledge this, as I might be the only person in a position to make this happen, and my time available to contribute to this patch set has been less than I would like.

But if it looks like we have all the pieces in place to base cpusets on containers, with no known regressions in cpuset capability, then we must find a way to ensure that one of these batch schedulers, using cpusets on a NUMA box, still works.

Signature · 3 lines
                  I won't rest till it's the best ...
                  Programmer, Linux Scalability
                  Paul Jackson <[email protected]> 1.925.600.0401
#6 lore raw

Re: [ckrm-tech] [PATCH 1/9] Containers (V9): Basic container framework

Quoted 37 lines · Paul Jackson

[[ I have bcc'd one or more batch scheduler experts on this post.
They will know who they are, and should be aware that they are
not listed in the public cc list of this message. - pj ]]

Quoted 4 lines · responding to Paul Menage's Container patch set on lkml,

Documentation/containers.txt line 264

Quoted 3 lines
Documentation/containers.txt new file +3 -0
+*** notify_on_release is disabled in the current patch set. It may be
+*** reactivated in a future patch in a less-intrusive manner
+

Won't this break user space tools for cpusets?

Yes - disabling notify_on_release would definitely break some important uses of cpusets. This feature must be reactivated somehow before I'll sign up for putting this patch set in the main line.

Actually, after I posted a few days ago in another lkml post:

	http://lkml.org/lkml/2007/4/29/66

that just the simplest cpuset command:
mount -t cpuset cpuset /dev/cpuset
mkdir /dev/cpuset/foo
echo 0 > /dev/cpuset/foo/mems

caused an immediate kernel deadlock (Srivatsa has proposed a fix), it is pretty clear that this container patch set is not getting the cpuset testing it will need for acceptance. That's partly my fault.

The batch scheduler folks, such as the variants of PBS, LSF and SGE are major user of cpusets on NUMA hardware.

This container based replacement for cpusets isn't ready for the main line until at least one of those schedulers can run through one of their test suites. I hesitate to even acknowledge this, as I might be the only person in a position to make this happen, and my time available to contribute to this patch set has been less than I would like.

Quoted 5 lines

But if it looks like we have all the pieces in place to base cpusets on containers, with no known regressions in cpuset capability, then we must find a way to ensure that one of these batch schedulers, using cpusets on a NUMA box, still works.

Would it be possible to extract those test cases and integrate them with a testing framework like LTP? Do you have any regression test suite for cpusets that can be made available publicly so that any changes to cpusets can be validated?

The reason I ask for the test suite is that I suspect that the container framework will evolve further and a reliable testing mechanism would be extremely useful.

Signature · 4 lines
	Warm Regards,
	Balbir Singh
	Linux Technology Center
	IBM, ISTL
#7 lore raw

Re: [ckrm-tech] [PATCH 1/9] Containers (V9): Basic container framework

Quoted 4 lines · Balbir

Would it be possible to extract those test cases and integrate them with a testing framework like LTP? Do you have any regression test suite for cpusets that can be made available publicly so that any changes to cpusets can be validated?

There are essentially two sorts of cpuset regression tests of interest.

I have one such test, and the batch scheduler developers have various tests of their batch schedulers.

1) Testing batch schedulers against cpusets:

I doubt that the batch scheduler developers would be able to extract a cpuset test from their tests, or be able to share it if they did. Their tests tend to be large tests of batch schedulers, and only incidentally test cpusets -- if we break cpusets, in sometimes even subtle ways that they happen to depend on, we break them.

Sometimes there is no way to guess exactly what sorts of changes will break their code; we'll just have to schedule at least one run through one or more of them that rely heavily on cpusets before a change as big as rebasing cpusets on containers is reasonably safe. This test cycle won't be all that easy, so I'd wait until we are pretty close to what we think should be taken into the mainline kernel.

I suppose I will have to be the one co-ordinating this test, as I am the only one I know with a presence in both camps.

Once this test is done, from then forward, if we break them, we'll just have to deal with it as we do now, when the breakage shows up well down stream from the main kernel tree, at the point that a major batch scheduler release runs into a major distribution release containing the breakage. There is no practical way that I can see, as an ongoing basis, to continue testing for such breakage with every minor change to cpuset related code in the kernel. Any breakage found this way is dealt with by changes in user level code.

Once again, I have bcc'd one or more developers of batch schedulers, so they can see what nonsense I am spouting about them now ;).

2) Testing cpusets with a specific test.

There I can do better. Attached is the cpuset regression test I use. It requires at least 4 cpus and 2 memory nodes to do anything useful. It is copyright by SGI, released under GPL license.

This regression test is the primary cpuset test upon which I relied during the development of cpusets, and continue to rely. Except for one subtle race condition in the test itself, it has not changed in the last two to three years.

This test requires no user level code not found in an ordinary distro. It does require the taskset and numactl commands, for the purposes of testing certain interactions with them. It assumes that there are not other cpusets currently setup in the system that happen to conflict with the ones it creates.

See further comments within the test script itself.

Signature · 3 lines
                  I won't rest till it's the best ...
                  Programmer, Linux Scalability
                  Paul Jackson <[email protected]> 1.925.600.0401
#8 lore raw

Re: [ckrm-tech] [PATCH 1/9] Containers (V9): Basic container framework

Quoted 32 lines · Paul Jackson

Balbir wrote:

1) Testing batch schedulers against cpusets:

I doubt that the batch scheduler developers would be able to extract a cpuset test from their tests, or be able to share it if they did. Their tests tend to be large tests of batch schedulers, and only incidentally test cpusets -- if we break cpusets, in sometimes even subtle ways that they happen to depend on, we break them.

Sometimes there is no way to guess exactly what sorts of changes will break their code; we'll just have to schedule at least one run through one or more of them that rely heavily on cpusets before a change as big as rebasing cpusets on containers is reasonably safe. This test cycle won't be all that easy, so I'd wait until we are pretty close to what we think should be taken into the mainline kernel.

I suppose I will have to be the one co-ordinating this test, as I am the only one I know with a presence in both camps.

Once this test is done, from then forward, if we break them, we'll just have to deal with it as we do now, when the breakage shows up well down stream from the main kernel tree, at the point that a major batch scheduler release runs into a major distribution release containing the breakage. There is no practical way that I can see, as an ongoing basis, to continue testing for such breakage with every minor change to cpuset related code in the kernel. Any breakage found this way is dealt with by changes in user level code.

Quoted 3 lines

Once again, I have bcc'd one or more developers of batch schedulers, so they can see what nonsense I am spouting about them now ;).

That sounds reasonable to me

Quoted 11 lines

2) Testing cpusets with a specific test.

There I can do better. Attached is the cpuset regression test I use. It requires at least 4 cpus and 2 memory nodes to do anything useful. It is copyright by SGI, released under GPL license.

This regression test is the primary cpuset test upon which I relied during the development of cpusets, and continue to rely. Except for one subtle race condition in the test itself, it has not changed in the last two to three years.

Quoted 8 lines

This test requires no user level code not found in an ordinary distro. It does require the taskset and numactl commands, for the purposes of testing certain interactions with them. It assumes that there are not other cpusets currently setup in the system that happen to conflict with the ones it creates.

See further comments within the test script itself.

Thanks for the script. Would you like to contribute this script to LTP for wider availability and testing?

Signature · 4 lines
	Warm Regards,
	Balbir Singh
	Linux Technology Center
	IBM, ISTL
#9 lore raw

Re: [ckrm-tech] [PATCH 1/9] Containers (V9): Basic container framework

Quoted 2 lines · Balbir

Thanks for the script. Would you like to contribute this script to LTP for wider availability and testing?

Sounds good - though I'm a tad lazy, and don't know how to contribute this to the LTP.

You're welcome to contribute it yourself, if that's easier for you, or to point me in the direction of some information on this.

Since it's GPL, you can republish it, under that license.

... Most likely however, it will take a little tweaking to make this test work well for LTP. Most automated test environments have some requirements on the behaviour of each test, and it would be surprising if this test happen already to conform to LTP's requirements.

Signature · 3 lines
                  I won't rest till it's the best ...
                  Programmer, Linux Scalability
                  Paul Jackson <[email protected]> 1.925.600.0401
#10 lore raw

Re: [ckrm-tech] [PATCH 1/9] Containers (V9): Basic container framework

Quoted 10 lines · Paul Jackson
Quoted 2 lines · Balbir

Thanks for the script. Would you like to contribute this script to LTP for wider availability and testing?

Sounds good - though I'm a tad lazy, and don't know how to contribute this to the LTP.

You're welcome to contribute it yourself, if that's easier for you, or to point me in the direction of some information on this.

Quoted 7 lines

Since it's GPL, you can republish it, under that license.

... Most likely however, it will take a little tweaking to make this test work well for LTP. Most automated test environments have some requirements on the behaviour of each test, and it would be surprising if this test happen already to conform to LTP's requirements.

I'll figure out more, once I am comfortable with the script and the setup environment, I'll try and get this into LTP with a GPL license.

Signature · 4 lines
	Thanks,
	Balbir Singh
	Linux Technology Center
	IBM, ISTL
#11 lore raw

PATCH 2/9 Containers (V9): Example CPU accounting subsystem

In reply to [email protected]

This example demonstrates how to use the generic container subsystem for a simple resource tracker that counts, for the processes in a container, the total CPU time used and the %CPU used in the last complete 10 second interval.

Portions contributed by Balbir Singh <[email protected]>

---
 include/linux/container_subsys.h |    6 +
 include/linux/cpu_acct.h         |   14 ++
 init/Kconfig                     |    7 +
 kernel/Makefile                  |    1 
 kernel/cpu_acct.c                |  185 +++++++++++++++++++++++++++++++++++++++
 kernel/sched.c                   |   14 ++
 6 files changed, 224 insertions(+), 3 deletions(-)
include/linux/container_subsys.h +6 -0
Index: container-2.6.21-rc7-mm1/include/linux/container_subsys.h
===================================================================
@@ -7,4 +7,10 @@
 
 /* */
 
+#ifdef CONFIG_CONTAINER_CPUACCT
+SUBSYS(cpuacct)
+#endif
+
+/* */
+
 /* */
include/linux/cpu_acct.h new file +14 -0
Index: container-2.6.21-rc7-mm1/include/linux/cpu_acct.h
===================================================================
@@ -0,0 +1,14 @@
+
+#ifndef _LINUX_CPU_ACCT_H
+#define _LINUX_CPU_ACCT_H
+
+#include <linux/container.h>
+#include <asm/cputime.h>
+
+#ifdef CONFIG_CONTAINER_CPUACCT
+extern void cpuacct_charge(struct task_struct *, cputime_t cputime);
+#else
+static void inline cpuacct_charge(struct task_struct *p, cputime_t cputime) {}
+#endif
+
+#endif
init/Kconfig +7 -0
Index: container-2.6.21-rc7-mm1/init/Kconfig
===================================================================
@@ -322,6 +322,13 @@ config SYSFS_DEPRECATED
 	  If you are using a distro that was released in 2006 or later,
 	  it should be safe to say N here.
 
+config CONTAINER_CPUACCT
+	bool "Simple CPU accounting container subsystem"
+	select CONTAINERS
+	help
+	  Provides a simple Resource Controller for monitoring the
+	  total CPU consumed by the tasks in a container
+
 config RELAY
 	bool "Kernel->user space relay support (formerly relayfs)"
 	help
kernel/Makefile +1 -0
Index: container-2.6.21-rc7-mm1/kernel/Makefile
===================================================================
@@ -38,6 +38,7 @@ obj-$(CONFIG_KEXEC) += kexec.o
 obj-$(CONFIG_COMPAT) += compat.o
 obj-$(CONFIG_CONTAINERS) += container.o
 obj-$(CONFIG_CPUSETS) += cpuset.o
+obj-$(CONFIG_CONTAINER_CPUACCT) += cpu_acct.o
 obj-$(CONFIG_IKCONFIG) += configs.o
 obj-$(CONFIG_STOP_MACHINE) += stop_machine.o
 obj-$(CONFIG_AUDIT) += audit.o auditfilter.o
kernel/cpu_acct.c new file +185 -0

185 lines not on this page · raw message

kernel/sched.c +11 -3
Index: container-2.6.21-rc7-mm1/kernel/sched.c
===================================================================
@@ -56,6 +56,7 @@
 #include <linux/kprobes.h>
 #include <linux/delayacct.h>
 #include <linux/reciprocal_div.h>
+#include <linux/cpu_acct.h>
 
 #include <asm/tlb.h>
 #include <asm/unistd.h>
@@ -3328,9 +3329,13 @@ void account_user_time(struct task_struc
 {
 	struct cpu_usage_stat *cpustat = &kstat_this_cpu.cpustat;
 	cputime64_t tmp;
+	struct rq *rq = this_rq();
 
 	p->utime = cputime_add(p->utime, cputime);
 
+	if (p != rq->idle)
+		cpuacct_charge(p, cputime);
+
 	/* Add user time to cpustat. */
 	tmp = cputime_to_cputime64(cputime);
 	if (TASK_NICE(p) > 0)
@@ -3360,9 +3365,10 @@ void account_system_time(struct task_str
 		cpustat->irq = cputime64_add(cpustat->irq, tmp);
 	else if (softirq_count())
 		cpustat->softirq = cputime64_add(cpustat->softirq, tmp);
-	else if (p != rq->idle)
+	else if (p != rq->idle) {
 		cpustat->system = cputime64_add(cpustat->system, tmp);
-	else if (atomic_read(&rq->nr_iowait) > 0)
+		cpuacct_charge(p, cputime);
+	} else if (atomic_read(&rq->nr_iowait) > 0)
 		cpustat->iowait = cputime64_add(cpustat->iowait, tmp);
 	else
 		cpustat->idle = cputime64_add(cpustat->idle, tmp);
@@ -3387,8 +3393,10 @@ void account_steal_time(struct task_stru
 			cpustat->iowait = cputime64_add(cpustat->iowait, tmp);
 		else
 			cpustat->idle = cputime64_add(cpustat->idle, tmp);
-	} else
+	} else {
 		cpustat->steal = cputime64_add(cpustat->steal, tmp);
+		cpuacct_charge(p, -tmp);
+	}
 }
 
 /*

--

#12 lore raw

Re: [PATCH 2/9] Containers (V9): Example CPU accounting subsystem

kernel/cpu_acct.c line 85

Quoted 10 lines · [email protected] wrote:
kernel/cpu_acct.c new file +10 -0
+
+/* Lazily update the load calculation if necessary. Called with ca locked */
+static void cpuusage_update(struct cpuacct *ca)
+{
+	u64 now = get_jiffies_64();
+	/* If we're not due for an update, return */
+	if (ca->next_interval_check > now)
+		return;
+
+	if (ca->next_interval_check <= (now - INTERVAL)) {

These two conditions seem a little confusing.

If ca->next_interval_check > (now - INTERVAL), the else part is executed, but if ca->next_interval_check > (now - INTERVAL) then ca->next_interval_check > now, which implies we return and never enter the else part. It's been quite sometime since I looked at this code, so I might have gotten it wrong. I see a load of 0% on my powerpc box. I think it is because last_interval_time is always 0, I'll debug further

Signature · 4 lines
	Warm Regards,
	Balbir Singh
	Linux Technology Center
	IBM, ISTL
#13 lore raw

PATCH 3/9 Containers (V9): Add tasks file interface

In reply to [email protected]

This patch adds the per-directory "tasks" file for containerfs mounts; this allows the user to determine which tasks are members of a container by reading a container's "tasks", and to move a task into a container by writing its pid to its "tasks".

---
 include/linux/container.h |    2 
 kernel/container.c        |  344 ++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 346 insertions(+)
include/linux/container.h +2 -0
Index: container-2.6.21-rc7-mm1/include/linux/container.h
===================================================================
@@ -128,6 +128,8 @@ int container_is_removed(const struct co
 
 int container_path(const struct container *cont, char *buf, int buflen);
 
+int container_task_count(const struct container *cont);
+
 /* Return true if the container is a descendant of the current container */
 int container_is_descendant(const struct container *cont);
 
kernel/container.c +344 -0

368 lines not on this page · raw message

--

#14 lore raw

Re: [PATCH 3/9] Containers (V9): Add tasks file interface

kernel/container.c static inline void get_first_subsys(cons · line 758

Quoted 11 lines
kernel/container.c +11 -0
+static int attach_task_by_pid(struct container *cont, char *pidbuf)
+{
+	pid_t pid;
+	struct task_struct *tsk;
+	int ret;
+
+	if (sscanf(pidbuf, "%d", &pid) != 1)
+		return -EIO;
+
+	if (pid) {
+		read_lock(&tasklist_lock);

You could just use rcu_read_lock() and rcu_read_unlock() instead of read_lock(&tasklist_lock) and read_unlock(&tasklist_lock).

Quoted 51 lines
kernel/container.c +44 -0
+
+		tsk = find_task_by_pid(pid);
+		if (!tsk || tsk->flags & PF_EXITING) {
+			read_unlock(&tasklist_lock);
+			return -ESRCH;
+		}
+
+		get_task_struct(tsk);
+		read_unlock(&tasklist_lock);
+
+		if ((current->euid) && (current->euid != tsk->uid)
+		    && (current->euid != tsk->suid)) {
+			put_task_struct(tsk);
+			return -EACCES;
+		}
+	} else {
+		tsk = current;
+		get_task_struct(tsk);
+	}
+
+	ret = attach_task(cont, tsk);
+	put_task_struct(tsk);
+	return ret;
+}
+
 /* The various types of files and directories in a container file system */
 
 typedef enum {
@@ -684,6 +789,54 @@ typedef enum {
 	FILE_TASKLIST,
 } container_filetype_t;
 
+static ssize_t container_common_file_write(struct container *cont,
+					   struct cftype *cft,
+					   struct file *file,
+					   const char __user *userbuf,
+					   size_t nbytes, loff_t *unused_ppos)
+{
+	container_filetype_t type = cft->private;
+	char *buffer;
+	int retval = 0;
+
+	if (nbytes >= PATH_MAX)
+		return -E2BIG;
+
+	/* +1 for nul-terminator */
+	if ((buffer = kmalloc(nbytes + 1, GFP_KERNEL)) == 0)
+		return -ENOMEM;
+
+	if (copy_from_user(buffer, userbuf, nbytes)) {
+		retval = -EFAULT;

kernel/container.c typedef enum { · line 820

Quoted 10 lines
kernel/container.c +10 -0
+		goto out1;
+	}
+	buffer[nbytes] = 0;	/* nul-terminate */
+
+	mutex_lock(&container_mutex);
+
+	if (container_is_removed(cont)) {
+		retval = -ENODEV;
+		goto out2;
+	}

Can't we make this check prior to kmalloc() and copy_from_user()?

kernel/container.c int container_add_files(struct container · line 1040

Quoted 8 lines
kernel/container.c +8 -0
+int container_task_count(const struct container *cont) {
+	int count = 0;
+	struct task_struct *g, *p;
+	struct container_subsys_state *css;
+	int subsys_id;
+	get_first_subsys(cont, &css, &subsys_id);
+
+	read_lock(&tasklist_lock);

Can be replaced with rcu_read_lock() and rcu_read_unlock()

Quoted 7 lines
kernel/container.c +7 -0
+	do_each_thread(g, p) {
+		if (task_subsys_state(p, subsys_id) == css)
+			count ++;
+	} while_each_thread(g, p);
+	read_unlock(&tasklist_lock);
+	return count;
+}

kernel/container.c int container_add_files(struct container · line 1086

Quoted 10 lines
kernel/container.c +10 -0
+
+static int pid_array_load(pid_t *pidarray, int npids, struct container *cont)
+{
+	int n = 0;
+	struct task_struct *g, *p;
+	struct container_subsys_state *css;
+	int subsys_id;
+	get_first_subsys(cont, &css, &subsys_id);
+	rcu_read_lock();
+	read_lock(&tasklist_lock);

The read_lock() and read_unlock() are redundant

Quoted 15 lines
kernel/container.c +15 -0
+
+	do_each_thread(g, p) {
+		if (task_subsys_state(p, subsys_id) == css) {
+			pidarray[n++] = pid_nr(task_pid(p));
+			if (unlikely(n == npids))
+				goto array_full;
+		}
+	} while_each_thread(g, p);
+
+array_full:
+	read_unlock(&tasklist_lock);
+	rcu_read_unlock();
+	return n;
+}
+

[snip]

Quoted 38 lines
kernel/container.c +38 -0
+static int container_tasks_open(struct inode *unused, struct file *file)
+{
+	struct container *cont = __d_cont(file->f_dentry->d_parent);
+	struct ctr_struct *ctr;
+	pid_t *pidarray;
+	int npids;
+	char c;
+
+	if (!(file->f_mode & FMODE_READ))
+		return 0;
+
+	ctr = kmalloc(sizeof(*ctr), GFP_KERNEL);
+	if (!ctr)
+		goto err0;
+
+	/*
+	 * If container gets more users after we read count, we won't have
+	 * enough space - tough.  This race is indistinguishable to the
+	 * caller from the case that the additional container users didn't
+	 * show up until sometime later on.
+	 */
+	npids = container_task_count(cont);
+	pidarray = kmalloc(npids * sizeof(pid_t), GFP_KERNEL);
+	if (!pidarray)
+		goto err1;
+
+	npids = pid_array_load(pidarray, npids, cont);
+	sort(pidarray, npids, sizeof(pid_t), cmppid, NULL);
+
+	/* Call pid_array_to_buf() twice, first just to get bufsz */
+	ctr->bufsz = pid_array_to_buf(&c, sizeof(c), pidarray, npids) + 1;
+	ctr->buf = kmalloc(ctr->bufsz, GFP_KERNEL);
+	if (!ctr->buf)
+		goto err2;
+	ctr->bufsz = pid_array_to_buf(ctr->buf, ctr->bufsz, pidarray, npids);
+
+	kfree(pidarray);
+	file->private_data = ctr;

kernel/container.c int container_add_files(struct container · line 1175

Quoted 10 lines
kernel/container.c +10 -0
+	return 0;
+
+err2:
+	kfree(pidarray);
+err1:
+	kfree(ctr);
+err0:
+	return -ENOMEM;
+}
+

Any chance we could get a per-container task list? It will help subsystem writers as well. Alternatively, subsystems could use the attach_task() callback to track all tasks, but a per-container list will avoid duplication.

Signature · 4 lines
	Warm Regards,
	Balbir Singh
	Linux Technology Center
	IBM, ISTL
#15 lore raw

Re: [ckrm-tech] [PATCH 3/9] Containers (V9): Add tasks file interface

Quoted 6 lines · Balbir Singh

kernel/container.c typedef enum { · line 820

Quoted 4 lines
kernel/container.c +4 -0
+     if (container_is_removed(cont)) {
+             retval = -ENODEV;
+             goto out2;
+     }

Can't we make this check prior to kmalloc() and copy_from_user()?

We could but I'm not sure what it would buy us - we'd be optimizing for the case that essentially never occurs.

Quoted 13 lines

kernel/container.c int container_add_files(struct container · line 1040

Quoted 11 lines
kernel/container.c +8 -0
+int container_task_count(const struct container *cont) {
+     int count = 0;
+     struct task_struct *g, *p;
+     struct container_subsys_state *css;
+     int subsys_id;
+     get_first_subsys(cont, &css, &subsys_id);
+
+     read_lock(&tasklist_lock);

Can be replaced with rcu_read_lock() and rcu_read_unlock()

Are you sure about that? I see many users of do_each_thread()/while_each_thread() taking a lock on tasklist_lock, and only one (fs/binfmt_elf.c) that's clearly relying on an RCU critical sections. Documentation?

Quoted 3 lines

Any chance we could get a per-container task list? It will help subsystem writers as well.

It would be possible, yes - but we probably wouldn't want the overhead (additional ref counts and list manipulations on every fork/exit) of it on by default. We could make it a config option that particular subsystems could select.

I guess the question is how useful is this really, compared to just doing a do_each_thread() and seeing which tasks are in the container? Certainly that's a non-trivial operation, but in what circumstances is it really necessary to do it?

Paul

#16 lore raw

Re: [ckrm-tech] [PATCH 3/9] Containers (V9): Add tasks file interface

Quoted 8 lines · Paul Menage
Quoted 2 lines

Any chance we could get a per-container task list? It will help subsystem writers as well.

It would be possible, yes - but we probably wouldn't want the overhead (additional ref counts and list manipulations on every fork/exit) of it on by default. We could make it a config option that particular subsystems could select.

Quoted 4 lines

I guess the question is how useful is this really, compared to just doing a do_each_thread() and seeing which tasks are in the container? Certainly that's a non-trivial operation, but in what circumstances is it really necessary to do it?

For the CPU controller I was working on, (a fast access to) such a list would have been valuable. Basically each task has a weight associated with it (p->load_weight) which is made to depend upon its class limit. Whenever the class limit changes, we need to go and change all its member task's ->load_weight value.

If you don't maintain the per-container task list, I guess I could still work around it, by either:

- Walk the task table and find relevant members, OR better perhaps
- Move p->load_weight to a class structure

Signature · 2 lines
Regards,
vatsa
#17 lore raw

Re: [ckrm-tech] [PATCH 3/9] Containers (V9): Add tasks file interface

Quoted 11 lines · Srivatsa Vaddagiri

For the CPU controller I was working on, (a fast access to) such a list would have been valuable. Basically each task has a weight associated with it (p->load_weight) which is made to depend upon its class limit. Whenever the class limit changes, we need to go and change all its member task's ->load_weight value.

If you don't maintain the per-container task list, I guess I could still work around it, by either:

- Walk the task table and find relevant members

That doesn't seem like a terrible solution to me, unless you expect the class limit to be changing incredibly frequently.

If we had multiple subsystems that needed to walk the container member list on a fast-path operation (e.g. to make a scheduling decision) that would be a good reason to maintain such a list.

Quoted 2 lines

perhaps
- Move p->load_weight to a class structure

Sounds like a good idea if you can do it - but if it's per-process, how would it fit in the class structure?

Paul

#18 lore raw

Re: [ckrm-tech] [PATCH 3/9] Containers (V9): Add tasks file interface

Quoted 4 lines · Paul Menage
Quoted 1 line

- Walk the task table and find relevant members

That doesn't seem like a terrible solution to me, unless you expect the class limit to be changing incredibly frequently.

yeah i agree. Group limit(s) should not be changing so frequently.

Quoted 5 lines
Quoted 2 lines

perhaps
- Move p->load_weight to a class structure

Sounds like a good idea if you can do it - but if it's per-process, how would it fit in the class structure?

p->load_weight essentially depends on two things:

- nice value or static priority (which is per process, already present in task_struct)
- class limit (which is per class)

So in theory we can eliminate the load_weight field in task_struct and compute it at runtime from the above two fields, although it will be slightly inefficient I guess to compute the value every time a task is added to the runqueue. If that is not desirable, then we can stick with option 1 (walk task list and change member task's->load_weight upon class limit change).

Signature · 2 lines
Regards,
vatsa
#19 lore raw

Re: [ckrm-tech] [PATCH 3/9] Containers (V9): Add tasks file interface

In reply to Paul Menage
Quoted 10 lines · Paul Menage
Quoted 9 lines · Srivatsa Vaddagiri

For the CPU controller I was working on, (a fast access to) such a list would have been valuable. Basically each task has a weight associated with it (p->load_weight) which is made to depend upon its class limit. Whenever the class limit changes, we need to go and change all its member task's ->load_weight value.

If you don't maintain the per-container task list, I guess I could still work around it, by either:

Quoted 9 lines
Quoted 1 line

- Walk the task table and find relevant members

That doesn't seem like a terrible solution to me, unless you expect the class limit to be changing incredibly frequently.

If we had multiple subsystems that needed to walk the container member list on a fast-path operation (e.g. to make a scheduling decision) that would be a good reason to maintain such a list.

I now have a use case for maintaining a per-container task list. I am trying to build a per-container stats similar to taskstats. I intend to support container accounting of

1. Tasks running
2. Tasks stopped
3. Tasks un-interruptible
4. Tasks blocked on IO
5. Tasks sleeping

This would provide statistics similar to the patch that Pavel had sent out.

I faced the following problems while trying to implement this feature

1. There is no easy way to get a list of all tasks belonging to a container (we need to walk all threads)
2. There is no concept of a container identifier. When a user issues a command to extract statistics, the only unique container identifier is the container path, which means that we need to do a path lookup to determine the dentry for the container (which gets quite ugly with all the string manipulation)

Adding a container id, will make it easier to find a container and return statistics belonging to the container.

If we get these two features added to the current patches, it will make the infrastructure more "developer" and eventually more user friendly :-)

Signature · 4 lines
	Warm Regards,
	Balbir Singh
	Linux Technology Center
	IBM, ISTL
#20 lore raw

Re: [ckrm-tech] [PATCH 3/9] Containers (V9): Add tasks file interface

Quoted 11 lines · Balbir Singh

I now have a use case for maintaining a per-container task list. I am trying to build a per-container stats similar to taskstats. I intend to support container accounting of

1. Tasks running
2. Tasks stopped
3. Tasks un-interruptible
4. Tasks blocked on IO
5. Tasks sleeping

Quoted 6 lines

This would provide statistics similar to the patch that Pavel had sent out.

I faced the following problems while trying to implement this feature

1. There is no easy way to get a list of all tasks belonging to a container (we need to walk all threads)

Well, walking the taks list is pretty easy - but yes, it could become inefficient when there are many small containers in use.

I've got some ideas for a way of tracking this specifically for containers with subsystems that want this, while avoiding the overhead for subsystems that don't really need it. I'll try to add them to the next patchset.

Quoted 4 lines

2. There is no concept of a container identifier. When a user issues a command to extract statistics, the only unique container identifier is the container path, which means that we need to do a path lookup to determine the dentry for the container (which gets quite ugly with all the string manipulation)

We could just cache the container path permanently in the container, and invalidate it if any of its parents gets renamed. (I imagine this happens almost never.)

Quoted 3 lines

Adding a container id, will make it easier to find a container and return statistics belonging to the container.

Not unreasonable, but there are a few questions that would have to be answered:

- how is the container id picked? Like a pid, or user-defined? Or some kind of string?

- how would it be exposed to userspace? A generic control file provided by the container filesystem in all container directories?

- can you give a more concrete example of how this would actually be useful? For your container stats, it seems that just reading a control file in the container's directory would give you the stats that you want, and userspace already knows the container's name/id since it opened the control file.

Paul

#21 lore raw

Re: [ckrm-tech] [PATCH 3/9] Containers (V9): Add tasks file interface

Quoted 21 lines · Paul Menage
Quoted 11 lines · Balbir Singh

I now have a use case for maintaining a per-container task list. I am trying to build a per-container stats similar to taskstats. I intend to support container accounting of

1. Tasks running
2. Tasks stopped
3. Tasks un-interruptible
4. Tasks blocked on IO
5. Tasks sleeping

Quoted 8 lines

This would provide statistics similar to the patch that Pavel had sent out.

I faced the following problems while trying to implement this feature

1. There is no easy way to get a list of all tasks belonging to a container
(we need to walk all threads)

Quoted 7 lines

Well, walking the taks list is pretty easy - but yes, it could become inefficient when there are many small containers in use.

I've got some ideas for a way of tracking this specifically for containers with subsystems that want this, while avoiding the overhead for subsystems that don't really need it. I'll try to add them to the next patchset.

Super!

Quoted 10 lines
Quoted 9 lines

2. There is no concept of a container identifier. When a user issues a command
to extract statistics, the only unique container identifier is the
container
path, which means that we need to do a path lookup to determine the
dentry
for the container (which gets quite ugly with all the string
manipulation)

Quoted 4 lines

We could just cache the container path permanently in the container, and invalidate it if any of its parents gets renamed. (I imagine this happens almost never.)

Here's what I have so far, I cache the mount point of the container and add the container path to it. I'm now stuck examining tasks, while walking through a bunch of tasks, there is no easy way of knowing the container path of the task without walking all subsystems and then extracting the containers absolute path.

Quoted 11 lines
Quoted 4 lines

Adding a container id, will make it easier to find a container and
return
statistics belonging to the container.

Not unreasonable, but there are a few questions that would have to be answered:

- how is the container id picked? Like a pid, or user-defined? Or some kind of string?

I was planning on using a hierarchical scheme, top 8 bits for the container hierarchy and bottom 24 for a unique id. The id is automatically selected. Once we know the container id, we'll need a more efficient mechanism to map the id to the container.

Quoted 3 lines

- how would it be exposed to userspace? A generic control file provided by the container filesystem in all container directories?

A file in all container directories is an option

Quoted 6 lines

- can you give a more concrete example of how this would actually be useful? For your container stats, it seems that just reading a control file in the container's directory would give you the stats that you want, and userspace already knows the container's name/id since it opened the control file.

Sure, the plan is to build a containerstats interface like taskstats. In taskstats, we exchange data between user space and kernel space using genetlink sockets. We have a push and pull mechanism for statistics.

Quoted 1 line

Paul

Signature · 4 lines
	Warm Regards,
	Balbir Singh
	Linux Technology Center
	IBM, ISTL
#22 lore raw

Re: [ckrm-tech] [PATCH 3/9] Containers (V9): Add tasks file interface

In reply to Paul Menage
Quoted 11 lines · Paul Menage
Quoted 6 lines · Balbir Singh

kernel/container.c typedef enum { · line 820

Quoted 4 lines
kernel/container.c +4 -0
+     if (container_is_removed(cont)) {
+             retval = -ENODEV;
+             goto out2;
+     }

Can't we make this check prior to kmalloc() and copy_from_user()?

We could but I'm not sure what it would buy us - we'd be optimizing for the case that essentially never occurs.

I am not sure about the never occurs part of it, because we check for the condition, so it could occur. I agree, it is a premature optimization and could wait a little longer before going in.

Quoted 14 lines
Quoted 13 lines

kernel/container.c int container_add_files(struct container · line 1040

Quoted 11 lines
kernel/container.c +8 -0
+int container_task_count(const struct container *cont) {
+     int count = 0;
+     struct task_struct *g, *p;
+     struct container_subsys_state *css;
+     int subsys_id;
+     get_first_subsys(cont, &css, &subsys_id);
+
+     read_lock(&tasklist_lock);

Can be replaced with rcu_read_lock() and rcu_read_unlock()

Quoted 5 lines

Are you sure about that? I see many users of do_each_thread()/while_each_thread() taking a lock on tasklist_lock, and only one (fs/binfmt_elf.c) that's clearly relying on an RCU critical sections. Documentation?

I suspect they are all pending conversions to be made. Eric is the expert on this. Meanwhile here's a couple of pointers. Quoting from the second URL

"We don't need the tasklist_lock to safely iterate through processes anymore."

http://www.linuxjournal.com/article/6993 (please see incremental use of RCU) and
http://kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.17/2.6.17-mm2/broken-out/proc-remove-tasklist_lock-from-proc_pid_readdir.patch

Quoted 15 lines
Quoted 3 lines

Any chance we could get a per-container task list? It will help subsystem writers as well.

It would be possible, yes - but we probably wouldn't want the overhead (additional ref counts and list manipulations on every fork/exit) of it on by default. We could make it a config option that particular subsystems could select.

I guess the question is how useful is this really, compared to just doing a do_each_thread() and seeing which tasks are in the container? Certainly that's a non-trivial operation, but in what circumstances is it really necessary to do it?

Paul

Signature · 4 lines
	Warm Regards,
	Balbir Singh
	Linux Technology Center
	IBM, ISTL
#23 lore raw

PATCH 4/9 Containers (V9): Add fork/exit hooks

In reply to [email protected]

This patch adds the necessary hooks to the fork() and exit() paths to ensure that new children inherit their parent's container assignments, and that exiting processes release reference counts on their containers.

---
 include/linux/container.h |    6 ++
 kernel/container.c        |  126 ++++++++++++++++++++++++++++++++++++++++++++++
 kernel/exit.c             |    2 
 kernel/fork.c             |   14 ++++-
 4 files changed, 146 insertions(+), 2 deletions(-)
kernel/exit.c +2 -0
Index: container-2.6.21-rc7-mm1/kernel/exit.c
===================================================================
@@ -32,6 +32,7 @@
 #include <linux/taskstats_kern.h>
 #include <linux/delayacct.h>
 #include <linux/cpuset.h>
+#include <linux/container.h>
 #include <linux/syscalls.h>
 #include <linux/signal.h>
 #include <linux/posix-timers.h>
@@ -939,6 +940,7 @@ fastcall NORET_TYPE void do_exit(long co
 	__exit_fs(tsk);
 	exit_thread();
 	cpuset_exit(tsk);
+	container_exit(tsk, 1);
 	exit_keys(tsk);
 
 	if (group_dead && tsk->signal->leader)
kernel/fork.c +12 -2
Index: container-2.6.21-rc7-mm1/kernel/fork.c
===================================================================
@@ -30,6 +30,7 @@
 #include <linux/capability.h>
 #include <linux/cpu.h>
 #include <linux/cpuset.h>
+#include <linux/container.h>
 #include <linux/security.h>
 #include <linux/swap.h>
 #include <linux/syscalls.h>
@@ -962,6 +963,7 @@ static struct task_struct *copy_process(
 {
 	int retval;
 	struct task_struct *p = NULL;
+	int container_callbacks_done = 0;
 
 	if ((clone_flags & (CLONE_NEWNS|CLONE_FS)) == (CLONE_NEWNS|CLONE_FS))
 		return ERR_PTR(-EINVAL);
@@ -1061,12 +1063,13 @@ static struct task_struct *copy_process(
 	p->io_wait = NULL;
 	p->audit_context = NULL;
 	cpuset_fork(p);
+	container_fork(p);
 #ifdef CONFIG_NUMA
  	p->mempolicy = mpol_copy(p->mempolicy);
  	if (IS_ERR(p->mempolicy)) {
  		retval = PTR_ERR(p->mempolicy);
  		p->mempolicy = NULL;
- 		goto bad_fork_cleanup_cpuset;
+ 		goto bad_fork_cleanup_container;
  	}
 	mpol_fix_fork_child_flag(p);
 #endif
@@ -1176,6 +1179,12 @@ static struct task_struct *copy_process(
 	/* Perform scheduler related setup. Assign this task to a CPU. */
 	sched_fork(p, clone_flags);
 
+	/* Now that the task is set up, run container callbacks if
+	 * necessary. We need to run them before the task is visible
+	 * on the tasklist. */
+	container_fork_callbacks(p);
+	container_callbacks_done = 1;
+
 	/* Need tasklist lock for parent etc handling! */
 	write_lock_irq(&tasklist_lock);
 
@@ -1298,9 +1307,10 @@ bad_fork_cleanup_security:
 bad_fork_cleanup_policy:
 #ifdef CONFIG_NUMA
 	mpol_free(p->mempolicy);
-bad_fork_cleanup_cpuset:
+bad_fork_cleanup_container:
 #endif
 	cpuset_exit(p);
+	container_exit(p, container_callbacks_done);
 	delayacct_tsk_free(p);
 	if (p->binfmt)
 		module_put(p->binfmt->module);
include/linux/container.h +6 -0
Index: container-2.6.21-rc7-mm1/include/linux/container.h
===================================================================
@@ -18,6 +18,9 @@
 extern int container_init_early(void);
 extern int container_init(void);
 extern void container_init_smp(void);
+extern void container_fork(struct task_struct *p);
+extern void container_fork_callbacks(struct task_struct *p);
+extern void container_exit(struct task_struct *p, int run_callbacks);
 
 extern struct file_operations proc_container_operations;
 
@@ -191,6 +194,9 @@ int container_path(const struct containe
 static inline int container_init_early(void) { return 0; }
 static inline int container_init(void) { return 0; }
 static inline void container_init_smp(void) {}
+static inline void container_fork(struct task_struct *p) {}
+static inline void container_fork_callbacks(struct task_struct *p) {}
+static inline void container_exit(struct task_struct *p, int callbacks) {}
 
 static inline void container_lock(void) {}
 static inline void container_unlock(void) {}
kernel/container.c +126 -0

135 lines not on this page · raw message

--

#24 lore raw

PATCH 5/9 Containers (V9): Add container_clone() interface

In reply to [email protected]

This patch adds support for container_clone(), a speculative interface to creating new containers intended to be used for systems such as namespace unsharing.

---
 include/linux/container.h |    2 
 kernel/container.c        |  121 ++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 123 insertions(+)
kernel/container.c +121 -0
Index: container-2.6.21-rc7-mm1/kernel/container.c
===================================================================
@@ -1619,3 +1619,124 @@ void container_exit(struct task_struct *
 	tsk->containers = init_task.containers;
 	task_unlock(tsk);
 }
+
+static atomic_t namecnt;
+static void get_unused_name(char *buf) {
+	sprintf(buf, "node%d", atomic_inc_return(&namecnt));
+}
+
+/**
+ * container_clone - duplicate the current container in the hierarchy
+ * that the given subsystem is attached to, and move this task into
+ * the new child
+ */
+int container_clone(struct task_struct *tsk, struct container_subsys *subsys)
+{
+	struct dentry *dentry;
+	int ret = 0;
+	char nodename[32];
+	struct container *parent, *child;
+	struct inode *inode;
+	struct css_group *cg;
+	struct containerfs_root *root;
+
+	/* We shouldn't be called by an unregistered subsystem */
+	BUG_ON(!subsys->active);
+
+	/* First figure out what hierarchy and container we're dealing
+	 * with, and pin them so we can drop container_mutex */
+	mutex_lock(&container_mutex);
+ again:
+	root = subsys->root;
+	if (root == &rootnode) {
+		printk(KERN_INFO
+		       "Not cloning container for unused subsystem %s\n",
+		       subsys->name);
+		mutex_unlock(&container_mutex);
+		return 0;
+	}
+	cg = &tsk->containers;
+	parent = task_container(tsk, subsys->subsys_id);
+	/* Pin the hierarchy */
+	atomic_inc(&parent->root->sb->s_active);
+
+	mutex_unlock(&container_mutex);
+
+	/* Now do the VFS work to create a container */
+	get_unused_name(nodename);
+	inode = parent->dentry->d_inode;
+
+	/* Hold the parent directory mutex across this operation to
+	 * stop anyone else deleting the new container */
+	mutex_lock(&inode->i_mutex);
+	dentry = container_get_dentry(parent->dentry, nodename);
+	if (IS_ERR(dentry)) {
+		printk(KERN_INFO
+		       "Couldn't allocate dentry for %s: %ld\n", nodename,
+		       PTR_ERR(dentry));
+		ret = PTR_ERR(dentry);
+		goto out_release;
+	}
+
+	/* Create the container directory, which also creates the container */
+	ret = vfs_mkdir(inode, dentry, S_IFDIR | 0755);
+	child = __d_cont(dentry);
+	dput(dentry);
+	if (ret) {
+		printk(KERN_INFO
+		       "Failed to create container %s: %d\n", nodename,
+		       ret);
+		goto out_release;
+	}
+
+	if (!child) {
+		printk(KERN_INFO
+		       "Couldn't find new container %s\n", nodename);
+		ret = -ENOMEM;
+		goto out_release;
+	}
+
+	/* The container now exists. Retake container_mutex and check
+	 * that we're still in the same state that we thought we
+	 * were. */
+	mutex_lock(&container_mutex);
+	if ((root != subsys->root) ||
+	    (parent != task_container(tsk, subsys->subsys_id))) {
+		/* Aargh, we raced ... */
+		mutex_unlock(&inode->i_mutex);
+
+		deactivate_super(parent->root->sb);
+		/* The container is still accessible in the VFS, but
+		 * we're not going to try to rmdir() it at this
+		 * point. */
+		printk(KERN_INFO
+		       "Race in container_clone() - leaking container %s\n",
+		       nodename);
+		goto again;
+	}
+
+	/* All seems fine. Finish by moving the task into the new container */
+	ret = attach_task(child, tsk);
+	mutex_unlock(&container_mutex);
+
+ out_release:
+	mutex_unlock(&inode->i_mutex);
+	deactivate_super(parent->root->sb);
+	return ret;
+}
+
+/* See if "cont" is a descendant of the current task's container in
+ * the appropriate hierarchy */
+
+int container_is_descendant(const struct container *cont) {
+	int ret;
+	struct container *target;
+	int subsys_id;
+	get_first_subsys(cont, NULL, &subsys_id);
+	target = task_container(current, subsys_id);
+	while (cont != target && cont!= cont->top_container) {
+		cont = cont->parent;
+	}
+	ret = (cont == target);
+	return ret;
+}
include/linux/container.h +2 -0
Index: container-2.6.21-rc7-mm1/include/linux/container.h
===================================================================
@@ -189,6 +189,8 @@ static inline struct container* task_con
 
 int container_path(const struct container *cont, char *buf, int buflen);
 
+int container_clone(struct task_struct *tsk, struct container_subsys *ss);
+
 #else /* !CONFIG_CONTAINERS */
 
 static inline int container_init_early(void) { return 0; }

--

#25 lore raw

PATCH 6/9 Containers (V9): Add procfs interface

In reply to [email protected]

This patch adds:

/proc/containers - general system info

/proc/*/container - per-task container membership info

---
 fs/proc/base.c     |    7 ++
 kernel/container.c |  128 +++++++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 135 insertions(+)
fs/proc/base.c +7 -0
Index: container-2.6.21-rc7-mm1/fs/proc/base.c
===================================================================
@@ -68,6 +68,7 @@
 #include <linux/security.h>
 #include <linux/ptrace.h>
 #include <linux/seccomp.h>
+#include <linux/container.h>
 #include <linux/cpuset.h>
 #include <linux/audit.h>
 #include <linux/poll.h>
@@ -1980,6 +1981,9 @@ static const struct pid_entry tgid_base_
 #ifdef CONFIG_CPUSETS
 	REG("cpuset",     S_IRUGO, cpuset),
 #endif
+#ifdef CONFIG_CONTAINERS
+	REG("container",  S_IRUGO, container),
+#endif
 	INF("oom_score",  S_IRUGO, oom_score),
 	REG("oom_adj",    S_IRUGO|S_IWUSR, oom_adjust),
 #ifdef CONFIG_AUDITSYSCALL
@@ -2270,6 +2274,9 @@ static const struct pid_entry tid_base_s
 #ifdef CONFIG_CPUSETS
 	REG("cpuset",    S_IRUGO, cpuset),
 #endif
+#ifdef CONFIG_CONTAINERS
+	REG("container",  S_IRUGO, container),
+#endif
 	INF("oom_score", S_IRUGO, oom_score),
 	REG("oom_adj",   S_IRUGO|S_IWUSR, oom_adjust),
 #ifdef CONFIG_AUDITSYSCALL
kernel/container.c +128 -0

150 lines not on this page · raw message

--

#26 lore raw

PATCH 7/9 Containers (V9): Make cpusets a client of containers

In reply to [email protected]

This patch removes the filesystem support logic from the cpusets system and makes cpusets a container subsystem

---
 Documentation/cpusets.txt        |   91 +--
 fs/proc/base.c                   |    4 
 include/linux/container_subsys.h |    6 
 include/linux/cpuset.h           |   12 
 include/linux/mempolicy.h        |   12 
 include/linux/sched.h            |    3 
 init/Kconfig                     |    6 
 kernel/cpuset.c                  | 1146 +++++----------------------------------
 kernel/exit.c                    |    2 
 kernel/fork.c                    |    3 
 mm/mempolicy.c                   |    2 
 11 files changed, 236 insertions(+), 1051 deletions(-)
Documentation/cpusets.txt +42 -49
Index: container-2.6.21-rc7-mm1/Documentation/cpusets.txt
===================================================================
@@ -7,6 +7,7 @@ Written by [email protected]
 Portions Copyright (c) 2004-2006 Silicon Graphics, Inc.
 Modified by Paul Jackson <[email protected]>
 Modified by Christoph Lameter <[email protected]>
+Modified by Paul Menage <[email protected]>
 
 CONTENTS:
 =========
@@ -16,10 +17,9 @@ CONTENTS:
   1.2 Why are cpusets needed ?
   1.3 How are cpusets implemented ?
   1.4 What are exclusive cpusets ?
-  1.5 What does notify_on_release do ?
-  1.6 What is memory_pressure ?
-  1.7 What is memory spread ?
-  1.8 How do I use cpusets ?
+  1.5 What is memory_pressure ?
+  1.6 What is memory spread ?
+  1.7 How do I use cpusets ?
 2. Usage Examples and Syntax
   2.1 Basic Usage
   2.2 Adding/removing cpus
@@ -43,18 +43,19 @@ hierarchy visible in a virtual file syst
 hooks, beyond what is already present, required to manage dynamic
 job placement on large systems.
 
-Each task has a pointer to a cpuset.  Multiple tasks may reference
-the same cpuset.  Requests by a task, using the sched_setaffinity(2)
-system call to include CPUs in its CPU affinity mask, and using the
-mbind(2) and set_mempolicy(2) system calls to include Memory Nodes
-in its memory policy, are both filtered through that tasks cpuset,
-filtering out any CPUs or Memory Nodes not in that cpuset.  The
-scheduler will not schedule a task on a CPU that is not allowed in
-its cpus_allowed vector, and the kernel page allocator will not
-allocate a page on a node that is not allowed in the requesting tasks
-mems_allowed vector.
+Cpusets use the generic container subsystem described in
+Documentation/container.txt.
 
-User level code may create and destroy cpusets by name in the cpuset
+Requests by a task, using the sched_setaffinity(2) system call to
+include CPUs in its CPU affinity mask, and using the mbind(2) and
+set_mempolicy(2) system calls to include Memory Nodes in its memory
+policy, are both filtered through that tasks cpuset, filtering out any
+CPUs or Memory Nodes not in that cpuset.  The scheduler will not
+schedule a task on a CPU that is not allowed in its cpus_allowed
+vector, and the kernel page allocator will not allocate a page on a
+node that is not allowed in the requesting tasks mems_allowed vector.
+
+User level code may create and destroy cpusets by name in the container
 virtual file system, manage the attributes and permissions of these
 cpusets and which CPUs and Memory Nodes are assigned to each cpuset,
 specify and query to which cpuset a task is assigned, and list the
@@ -114,7 +115,7 @@ Cpusets extends these two mechanisms as 
  - Cpusets are sets of allowed CPUs and Memory Nodes, known to the
    kernel.
  - Each task in the system is attached to a cpuset, via a pointer
-   in the task structure to a reference counted cpuset structure.
+   in the task structure to a reference counted container structure.
  - Calls to sched_setaffinity are filtered to just those CPUs
    allowed in that tasks cpuset.
  - Calls to mbind and set_mempolicy are filtered to just
@@ -144,15 +145,10 @@ into the rest of the kernel, none in per
  - in page_alloc.c, to restrict memory to allowed nodes.
  - in vmscan.c, to restrict page recovery to the current cpuset.
 
-In addition a new file system, of type "cpuset" may be mounted,
-typically at /dev/cpuset, to enable browsing and modifying the cpusets
-presently known to the kernel.  No new system calls are added for
-cpusets - all support for querying and modifying cpusets is via
-this cpuset file system.
-
-Each task under /proc has an added file named 'cpuset', displaying
-the cpuset name, as the path relative to the root of the cpuset file
-system.
+You should mount the "container" filesystem type in order to enable
+browsing and modifying the cpusets presently known to the kernel.  No
+new system calls are added for cpusets - all support for querying and
+modifying cpusets is via this cpuset file system.
 
 The /proc/<pid>/status file for each task has two added lines,
 displaying the tasks cpus_allowed (on which CPUs it may be scheduled)
@@ -162,16 +158,15 @@ in the format seen in the following exam
   Cpus_allowed:   ffffffff,ffffffff,ffffffff,ffffffff
   Mems_allowed:   ffffffff,ffffffff
 
-Each cpuset is represented by a directory in the cpuset file system
-containing the following files describing that cpuset:
+Each cpuset is represented by a directory in the container file system
+containing (on top of the standard container files) the following
+files describing that cpuset:
 
  - cpus: list of CPUs in that cpuset
  - mems: list of Memory Nodes in that cpuset
  - memory_migrate flag: if set, move pages to cpusets nodes
  - cpu_exclusive flag: is cpu placement exclusive?
  - mem_exclusive flag: is memory placement exclusive?
- - tasks: list of tasks (by pid) attached to that cpuset
- - notify_on_release flag: run /sbin/cpuset_release_agent on exit?
  - memory_pressure: measure of how much paging pressure in cpuset
 
 In addition, the root cpuset only has the following file:
@@ -236,21 +231,7 @@ such as requests from interrupt handlers
 outside even a mem_exclusive cpuset.
 
 
-1.5 What does notify_on_release do ?
-------------------------------------
-
-If the notify_on_release flag is enabled (1) in a cpuset, then whenever
-the last task in the cpuset leaves (exits or attaches to some other
-cpuset) and the last child cpuset of that cpuset is removed, then
-the kernel runs the command /sbin/cpuset_release_agent, supplying the
-pathname (relative to the mount point of the cpuset file system) of the
-abandoned cpuset.  This enables automatic removal of abandoned cpusets.
-The default value of notify_on_release in the root cpuset at system
-boot is disabled (0).  The default value of other cpusets at creation
-is the current value of their parents notify_on_release setting.
-
-
-1.6 What is memory_pressure ?
+1.5 What is memory_pressure ?
 -----------------------------
 The memory_pressure of a cpuset provides a simple per-cpuset metric
 of the rate that the tasks in a cpuset are attempting to free up in
@@ -307,7 +288,7 @@ the tasks in the cpuset, in units of rec
 times 1000.
 
 
-1.7 What is memory spread ?
+1.6 What is memory spread ?
 ---------------------------
 There are two boolean flag files per cpuset that control where the
 kernel allocates pages for the file system buffers and related in
@@ -378,7 +359,7 @@ data set, the memory allocation across t
 can become very uneven.
 
 
-1.8 How do I use cpusets ?
+1.7 How do I use cpusets ?
 --------------------------
 
 In order to minimize the impact of cpusets on critical kernel
@@ -468,7 +449,7 @@ than stress the kernel.
 To start a new job that is to be contained within a cpuset, the steps are:
 
  1) mkdir /dev/cpuset
- 2) mount -t cpuset none /dev/cpuset
+ 2) mount -t container -ocpuset cpuset /dev/cpuset
  3) Create the new cpuset by doing mkdir's and write's (or echo's) in
     the /dev/cpuset virtual file system.
  4) Start a task that will be the "founding father" of the new job.
@@ -480,7 +461,7 @@ For example, the following sequence of c
 named "Charlie", containing just CPUs 2 and 3, and Memory Node 1,
 and then start a subshell 'sh' in that cpuset:
 
-  mount -t cpuset none /dev/cpuset
+  mount -t container -ocpuset cpuset /dev/cpuset
   cd /dev/cpuset
   mkdir Charlie
   cd Charlie
@@ -512,7 +493,7 @@ Creating, modifying, using the cpusets c
 virtual filesystem.
 
 To mount it, type:
-# mount -t cpuset none /dev/cpuset
+# mount -t container -o cpuset cpuset /dev/cpuset
 
 Then under /dev/cpuset you can find a tree that corresponds to the
 tree of the cpusets in the system. For instance, /dev/cpuset
@@ -555,6 +536,18 @@ To remove a cpuset, just use rmdir:
 This will fail if the cpuset is in use (has cpusets inside, or has
 processes attached).
 
+Note that for legacy reasons, the "cpuset" filesystem exists as a
+wrapper around the container filesystem.
+
+The command
+
+mount -t cpuset X /dev/cpuset
+
+is equivalent to
+
+mount -t container -ocpuset X /dev/cpuset
+echo "/sbin/cpuset_release_agent" > /dev/cpuset/release_agent
+
 2.2 Adding/removing cpus
 ------------------------
 
include/linux/cpuset.h +8 -4
Index: container-2.6.21-rc7-mm1/include/linux/cpuset.h
===================================================================
@@ -11,6 +11,7 @@
 #include <linux/sched.h>
 #include <linux/cpumask.h>
 #include <linux/nodemask.h>
+#include <linux/container.h>
 
 #ifdef CONFIG_CPUSETS
 
@@ -19,8 +20,6 @@ extern int number_of_cpusets;	/* How man
 extern int cpuset_init_early(void);
 extern int cpuset_init(void);
 extern void cpuset_init_smp(void);
-extern void cpuset_fork(struct task_struct *p);
-extern void cpuset_exit(struct task_struct *p);
 extern cpumask_t cpuset_cpus_allowed(struct task_struct *p);
 extern nodemask_t cpuset_mems_allowed(struct task_struct *p);
 #define cpuset_current_mems_allowed (current->mems_allowed)
@@ -75,13 +74,13 @@ static inline int cpuset_do_slab_mem_spr
 
 extern void cpuset_track_online_nodes(void);
 
+extern int current_cpuset_is_being_rebound(void);
+
 #else /* !CONFIG_CPUSETS */
 
 static inline int cpuset_init_early(void) { return 0; }
 static inline int cpuset_init(void) { return 0; }
 static inline void cpuset_init_smp(void) {}
-static inline void cpuset_fork(struct task_struct *p) {}
-static inline void cpuset_exit(struct task_struct *p) {}
 
 static inline cpumask_t cpuset_cpus_allowed(struct task_struct *p)
 {
@@ -146,6 +145,11 @@ static inline int cpuset_do_slab_mem_spr
 
 static inline void cpuset_track_online_nodes(void) {}
 
+static inline int current_cpuset_is_being_rebound(void)
+{
+	return 0;
+}
+
 #endif /* !CONFIG_CPUSETS */
 
 #endif /* _LINUX_CPUSET_H */
include/linux/mempolicy.h +0 -12
Index: container-2.6.21-rc7-mm1/include/linux/mempolicy.h
===================================================================
@@ -148,14 +148,6 @@ extern void mpol_rebind_task(struct task
 					const nodemask_t *new);
 extern void mpol_rebind_mm(struct mm_struct *mm, nodemask_t *new);
 extern void mpol_fix_fork_child_flag(struct task_struct *p);
-#define set_cpuset_being_rebound(x) (cpuset_being_rebound = (x))
-
-#ifdef CONFIG_CPUSETS
-#define current_cpuset_is_being_rebound() \
-				(cpuset_being_rebound == current->cpuset)
-#else
-#define current_cpuset_is_being_rebound() 0
-#endif
 
 extern struct mempolicy default_policy;
 extern struct zonelist *huge_zonelist(struct vm_area_struct *vma,
@@ -173,8 +165,6 @@ static inline void check_highest_zone(en
 int do_migrate_pages(struct mm_struct *mm,
 	const nodemask_t *from_nodes, const nodemask_t *to_nodes, int flags);
 
-extern void *cpuset_being_rebound;	/* Trigger mpol_copy vma rebind */
-
 #else
 
 struct mempolicy {};
@@ -253,8 +243,6 @@ static inline void mpol_fix_fork_child_f
 {
 }
 
-#define set_cpuset_being_rebound(x) do {} while (0)
-
 static inline struct zonelist *huge_zonelist(struct vm_area_struct *vma,
 		unsigned long addr, gfp_t gfp_flags)
 {
include/linux/sched.h +0 -3
Index: container-2.6.21-rc7-mm1/include/linux/sched.h
===================================================================
@@ -772,8 +772,6 @@ static inline int above_background_load(
 }
 
 struct io_context;			/* See blkdev.h */
-struct cpuset;
-
 #define NGROUPS_SMALL		32
 #define NGROUPS_PER_BLOCK	((int)(PAGE_SIZE / sizeof(gid_t)))
 struct group_info {
@@ -1095,7 +1093,6 @@ struct task_struct {
 	short il_next;
 #endif
 #ifdef CONFIG_CPUSETS
-	struct cpuset *cpuset;
 	nodemask_t mems_allowed;
 	int cpuset_mems_generation;
 	int cpuset_mem_spread_rotor;
init/Kconfig +6 -0
Index: container-2.6.21-rc7-mm1/init/Kconfig
===================================================================
@@ -294,6 +294,7 @@ config CONTAINERS
 config CPUSETS
 	bool "Cpuset support"
 	depends on SMP
+	select CONTAINERS
 	help
 	  This option will let you create and manage CPUSETs which
 	  allow dynamically partitioning a system into sets of CPUs and
@@ -329,6 +330,11 @@ config CONTAINER_CPUACCT
 	  Provides a simple Resource Controller for monitoring the
 	  total CPU consumed by the tasks in a container
 
+config PROC_PID_CPUSET
+	bool "Include legacy /proc/<pid>/cpuset file"
+	depends on CPUSETS
+	default y
+
 config RELAY
 	bool "Kernel->user space relay support (formerly relayfs)"
 	help
kernel/cpuset.c +172 -974

1,623 lines not on this page · raw message

kernel/exit.c +0 -2
Index: container-2.6.21-rc7-mm1/kernel/exit.c
===================================================================
@@ -31,7 +31,6 @@
 #include <linux/mempolicy.h>
 #include <linux/taskstats_kern.h>
 #include <linux/delayacct.h>
-#include <linux/cpuset.h>
 #include <linux/container.h>
 #include <linux/syscalls.h>
 #include <linux/signal.h>
@@ -939,7 +938,6 @@ fastcall NORET_TYPE void do_exit(long co
 	__exit_files(tsk);
 	__exit_fs(tsk);
 	exit_thread();
-	cpuset_exit(tsk);
 	container_exit(tsk, 1);
 	exit_keys(tsk);
 
kernel/fork.c +0 -3
Index: container-2.6.21-rc7-mm1/kernel/fork.c
===================================================================
@@ -29,7 +29,6 @@
 #include <linux/nsproxy.h>
 #include <linux/capability.h>
 #include <linux/cpu.h>
-#include <linux/cpuset.h>
 #include <linux/container.h>
 #include <linux/security.h>
 #include <linux/swap.h>
@@ -1062,7 +1061,6 @@ static struct task_struct *copy_process(
 	p->io_context = NULL;
 	p->io_wait = NULL;
 	p->audit_context = NULL;
-	cpuset_fork(p);
 	container_fork(p);
 #ifdef CONFIG_NUMA
  	p->mempolicy = mpol_copy(p->mempolicy);
@@ -1309,7 +1307,6 @@ bad_fork_cleanup_policy:
 	mpol_free(p->mempolicy);
 bad_fork_cleanup_container:
 #endif
-	cpuset_exit(p);
 	container_exit(p, container_callbacks_done);
 	delayacct_tsk_free(p);
 	if (p->binfmt)
mm/mempolicy.c +0 -2
Index: container-2.6.21-rc7-mm1/mm/mempolicy.c
===================================================================
@@ -1310,7 +1310,6 @@ EXPORT_SYMBOL(alloc_pages_current);
  * keeps mempolicies cpuset relative after its cpuset moves.  See
  * further kernel/cpuset.c update_nodemask().
  */
-void *cpuset_being_rebound;
 
 /* Slow path of a mempolicy copy */
 struct mempolicy *__mpol_copy(struct mempolicy *old)
@@ -1909,4 +1908,3 @@ out:
 		m->version = (vma != priv->tail_vma) ? vma->vm_start : 0;
 	return 0;
 }
-
fs/proc/base.c +2 -2
Index: container-2.6.21-rc7-mm1/fs/proc/base.c
===================================================================
@@ -1978,7 +1978,7 @@ static const struct pid_entry tgid_base_
 #ifdef CONFIG_SCHEDSTATS
 	INF("schedstat",  S_IRUGO, pid_schedstat),
 #endif
-#ifdef CONFIG_CPUSETS
+#ifdef CONFIG_PROC_PID_CPUSET
 	REG("cpuset",     S_IRUGO, cpuset),
 #endif
 #ifdef CONFIG_CONTAINERS
@@ -2271,7 +2271,7 @@ static const struct pid_entry tid_base_s
 #ifdef CONFIG_SCHEDSTATS
 	INF("schedstat", S_IRUGO, pid_schedstat),
 #endif
-#ifdef CONFIG_CPUSETS
+#ifdef CONFIG_PROC_PID_CPUSET
 	REG("cpuset",    S_IRUGO, cpuset),
 #endif
 #ifdef CONFIG_CONTAINERS
include/linux/container_subsys.h +6 -0
Index: container-2.6.21-rc7-mm1/include/linux/container_subsys.h
===================================================================
@@ -13,4 +13,10 @@ SUBSYS(cpuacct)
 
 /* */
 
+#ifdef CONFIG_CPUSETS
+SUBSYS(cpuset)
+#endif
+
+/* */
+
 /* */

--

#27 lore raw

PATCH 8/9 Containers (V9): Share css_group arrays between tasks with same container memberships

In reply to [email protected]

This patch replaces the struct css_group embedded in task_struct with a pointer; all tasks with the same set of memberships across all hierarchies will share a css_group object.

The css_group used by init isn't refcounted, since it can't ever be freed; this speeds up fork/exit for any systems that have containers compiled in but haven't actually created any containers other than the default one.

With more than one registered subsystem, this reduces the number of atomic inc/dec operations required when tasks fork/exit;

Assuming that many tasks share the same container assignments, this reduces overall space usage and keeps the size of the task_struct down (only one pointer added to task_struct compared to a non-containers kernel).

---
 include/linux/container.h |   35 ++++++
 include/linux/sched.h     |   30 -----
 kernel/container.c        |  248 +++++++++++++++++++++++++++++++++++++---------
 3 files changed, 238 insertions(+), 75 deletions(-)
include/linux/container.h +34 -1
Index: container-2.6.21-rc7-mm1/include/linux/container.h
===================================================================
@@ -29,6 +29,14 @@ extern void container_unlock(void);
 
 struct containerfs_root;
 
+/* Define the enumeration of all container subsystems */
+#define SUBSYS(_x) _x ## _subsys_id,
+enum container_subsys_id {
+#include <linux/container_subsys.h>
+	CONTAINER_SUBSYS_COUNT
+};
+#undef SUBSYS
+
 /* Per-subsystem/per-container state maintained by the system. */
 struct container_subsys_state {
 	/* The container that this subsystem is attached to. Useful
@@ -87,6 +95,31 @@ struct container {
 	struct container *top_container;
 };
 
+/* A css_group is a structure holding pointers to a set of
+ * container_subsys_state objects. This saves space in the task struct
+ * object and speeds up fork()/exit(), since a single inc/dec can bump
+ * the reference count on the entire container set for a task.
+ */
+
+struct css_group {
+
+	/* Reference count */
+	struct kref ref;
+
+	/* List running through all container groups */
+	struct list_head list;
+
+	/* Set of subsystem states, one for each subsystem. NULL for
+	 * subsystems that aren't part of this hierarchy. These
+	 * pointers reduce the number of dereferences required to get
+	 * from a task to its state for a given container, but result
+	 * in increased space usage if tasks are in wildly different
+	 * groupings across different hierarchies. This array is
+	 * immutable after creation */
+	struct container_subsys_state *subsys[CONTAINER_SUBSYS_COUNT];
+
+};
+
 /* struct cftype:
  *
  * The files in the container filesystem mostly have a very simple read/write
@@ -178,7 +211,7 @@ static inline struct container_subsys_st
 static inline struct container_subsys_state *task_subsys_state(
 	struct task_struct *task, int subsys_id)
 {
-	return rcu_dereference(task->containers.subsys[subsys_id]);
+	return rcu_dereference(task->containers->subsys[subsys_id]);
 }
 
 static inline struct container* task_container(struct task_struct *task,
include/linux/sched.h +1 -29
Index: container-2.6.21-rc7-mm1/include/linux/sched.h
===================================================================
@@ -818,34 +818,6 @@ struct uts_namespace;
 
 struct prio_array;
 
-#ifdef CONFIG_CONTAINERS
-
-#define SUBSYS(_x) _x ## _subsys_id,
-enum container_subsys_id {
-#include <linux/container_subsys.h>
-	CONTAINER_SUBSYS_COUNT
-};
-#undef SUBSYS
-
-/* A css_group is a structure holding pointers to a set of
- * container_subsys_state objects.
- */
-
-struct css_group {
-
-	/* Set of subsystem states, one for each subsystem. NULL for
-	 * subsystems that aren't part of this hierarchy. These
-	 * pointers reduce the number of dereferences required to get
-	 * from a task to its state for a given container, but result
-	 * in increased space usage if tasks are in wildly different
-	 * groupings across different hierarchies. This array is
-	 * immutable after creation */
-	struct container_subsys_state *subsys[CONTAINER_SUBSYS_COUNT];
-
-};
-
-#endif /* CONFIG_CONTAINERS */
-
 struct task_struct {
 	volatile long state;	/* -1 unrunnable, 0 runnable, >0 stopped */
 	struct thread_info *thread_info;
@@ -1098,7 +1070,7 @@ struct task_struct {
 	int cpuset_mem_spread_rotor;
 #endif
 #ifdef CONFIG_CONTAINERS
-	struct css_group containers;
+	struct css_group *containers;
 #endif
 	struct robust_list_head __user *robust_list;
 #ifdef CONFIG_COMPAT
kernel/container.c +203 -45

364 lines not on this page · raw message

--

#28 lore raw

PATCH 9/9 Containers (V9): Simple debug info subsystem

In reply to [email protected]

This example subsystem exports debugging information as an aid to diagnosing refcount leaks, etc, in the container framework.

---
 include/linux/container_subsys.h |    4 +
 init/Kconfig                     |   10 ++++
 kernel/Makefile                  |    1 
 kernel/container_debug.c         |   89 +++++++++++++++++++++++++++++++++++++++
 4 files changed, 104 insertions(+)
include/linux/container_subsys.h +4 -0
Index: container-2.6.21-rc7-mm1/include/linux/container_subsys.h
===================================================================
@@ -19,4 +19,8 @@ SUBSYS(cpuset)
 
 /* */
 
+#ifdef CONFIG_CONTAINER_DEBUG
+SUBSYS(debug)
+#endif
+
 /* */
init/Kconfig +10 -0
Index: container-2.6.21-rc7-mm1/init/Kconfig
===================================================================
@@ -291,6 +291,16 @@ config IKCONFIG_PROC
 config CONTAINERS
 	bool
 
+config CONTAINER_DEBUG
+	bool "Example debug container subsystem"
+	select CONTAINERS
+	help
+	  This option enables a simple container subsystem that
+	  exports useful debugging information about the containers
+	  framework
+
+	  Say N if unsure
+
 config CPUSETS
 	bool "Cpuset support"
 	depends on SMP
kernel/container_debug.c new file +89 -0

89 lines not on this page · raw message

kernel/Makefile +1 -0
Index: container-2.6.21-rc7-mm1/kernel/Makefile
===================================================================
@@ -37,6 +37,7 @@ obj-$(CONFIG_BSD_PROCESS_ACCT) += acct.o
 obj-$(CONFIG_KEXEC) += kexec.o
 obj-$(CONFIG_COMPAT) += compat.o
 obj-$(CONFIG_CONTAINERS) += container.o
+obj-$(CONFIG_CONTAINER_DEBUG) += container_debug.o
 obj-$(CONFIG_CPUSETS) += cpuset.o
 obj-$(CONFIG_CONTAINER_CPUACCT) += cpu_acct.o
 obj-$(CONFIG_IKCONFIG) += configs.o

--

#29 lore raw
In reply to [email protected]
Quoted 3 lines

- decide whether "Containers" is an acceptable name for the system given its usage by some other development groups, or whether something else (ProcessSets? ResourceGroups? TaskGroups?) would be better

I place in nomination:

tasksets

However this would conflict with the taskset utility in Robert Loves schedutils package. The 'taskset' command "is used to set or retrieve the CPU affinity of a running process given its PID or to launch a new COMMAND with a given CPU affinity."

The various alternatives you listed all have the advantage of stating both that we have a container/group/set/collection/... of something, and -what- that something might be - process/resource/task/...

"tasksets" is the shortest spelling I could think of for such a compond "Something-Collection" form.

If the conflict with the schedutils utility is a concern, then the next best alternative would be:

taskgroups

This is the next shortest way to spell this "Something-Collection" form.

Signature · 3 lines
                  I won't rest till it's the best ...
                  Programmer, Linux Scalability
                  Paul Jackson <[email protected]> 1.925.600.0401
#30 lore raw
In reply to [email protected]

I'm afraid that this patch set doesn't do cpusets very well.

It builds and boots and mounts the cpuset file system ok. But trying to write the 'mems' file hangs the system hard.

I had to test it against 2.6.21-rc7-mm2, because I can't boot 2.6.21-rc7-mm1, due to the 'bad_page' problem that I noted in an earlier post this evening on lkml.

These container patches seemed to apply ok against 2.6.21-rc7-mm2, and built and booted ok. I built with an sn2_defconfig configuration, having the following CONTAINER and CPUSET settings:

CONFIG_CONTAINERS=y
CONFIG_CONTAINER_DEBUG=y
CONFIG_CPUSETS=y
CONFIG_CONTAINER_CPUACCT=y
CONFIG_PROC_PID_CPUSET=y
# CONFIG_ACPI_CONTAINER is not set

I could mount the cpuset file system on /dev/cpuset just fine.

Then I invoked the following commands:

    # cd /dev/cpuset
    # mkdir foo
    # cd foo
    # echo 0-3 > cpus
    # echo 0-1 > mems

At that point, the system hangs. Reproduced three times, on two boots. I never get a shell prompt back from that second echo. I have to hit Reset. The three different hangs were done with the following three different values:

	echo 0-3 > mems
	echo 0-1 > mems
	echo 1 > mems

On that last one, "echo 1 > mems", I did not do the echo to cpus first.

The test system had 8 cpus, numbered 0-7, and 4 mems, numbered 0-3.

Signature · 3 lines
                  I won't rest till it's the best ...
                  Programmer, Linux Scalability
                  Paul Jackson <[email protected]> 1.925.600.0401
#31 lore raw
Quoted 2 lines · Paul Jackson

It builds and boots and mounts the cpuset file system ok. But trying to write the 'mems' file hangs the system hard.

Basically we are attempting a read_lock(&tasklist_lock) in container_task_count() after taking write_lock_irq(&tasklist_lock) in update_nodemask()!

This patch seems to fix the prb for me:

Fix write_lock() followed by read_lock() bug by introducing a 2nd argument to be passed into container_task_count. Other choice is to introduce a lock and unlocked versions of container_task_count() ..

---

include/linux/container.h renamed from include/linux/container.h~lock_fix +1 -1
diff -puN include/linux/container.h~lock_fix include/linux/container.h
@@ -164,7 +164,7 @@ int container_is_removed(const struct co
 
 int container_path(const struct container *cont, char *buf, int buflen);
 
-int container_task_count(const struct container *cont);
+int container_task_count(const struct container *cont, int take_lock);
 
 /* Return true if the container is a descendant of the current container */
 int container_is_descendant(const struct container *cont);
kernel/container.c renamed from kernel/container.c~lock_fix +9 -5
diff -puN kernel/container.c~lock_fix kernel/container.c
@@ -1193,21 +1193,25 @@ int container_add_files(struct container
 /* Count the number of tasks in a container. Could be made more
  * time-efficient but less space-efficient with more linked lists
  * running through each container and the css_group structures
- * that referenced it. */
+ * that referenced it. 'take_lock' indicates whether caller has locked
+ * tasklist or not.
+ */
 
-int container_task_count(const struct container *cont) {
+int container_task_count(const struct container *cont, int take_lock) {
 	int count = 0;
 	struct task_struct *g, *p;
 	struct container_subsys_state *css;
 	int subsys_id;
 	get_first_subsys(cont, &css, &subsys_id);
 
-	read_lock(&tasklist_lock);
+	if (take_lock)
+		read_lock(&tasklist_lock);
 	do_each_thread(g, p) {
 		if (task_subsys_state(p, subsys_id) == css)
 			count ++;
 	} while_each_thread(g, p);
-	read_unlock(&tasklist_lock);
+	if (take_lock)
+		read_unlock(&tasklist_lock);
 	return count;
 }
 
@@ -1311,7 +1315,7 @@ static int container_tasks_open(struct i
 	 * caller from the case that the additional container users didn't
 	 * show up until sometime later on.
 	 */
-	npids = container_task_count(cont);
+	npids = container_task_count(cont, 1);
 	pidarray = kmalloc(npids * sizeof(pid_t), GFP_KERNEL);
 	if (!pidarray)
 		goto err1;
kernel/cpuset.c renamed from kernel/cpuset.c~lock_fix +2 -2
diff -puN kernel/cpuset.c~lock_fix kernel/cpuset.c
@@ -631,13 +631,13 @@ static int update_nodemask(struct cpuset
 	 * enough mmarray[] w/o using GFP_ATOMIC.
 	 */
 	while (1) {
-		ntasks = container_task_count(cs->css.container);  /* guess */
+		ntasks = container_task_count(cs->css.container, 1); /* guess */
 		ntasks += fudge;
 		mmarray = kmalloc(ntasks * sizeof(*mmarray), GFP_KERNEL);
 		if (!mmarray)
 			goto done;
 		write_lock_irq(&tasklist_lock);		/* block fork */
-		if (container_task_count(cs->css.container) <= ntasks)
+		if (container_task_count(cs->css.container, 0) <= ntasks)
 			break;				/* got enough */
 		write_unlock_irq(&tasklist_lock);	/* try again */
 		kfree(mmarray);
kernel/container_debug.c renamed from kernel/container_debug.c~lock_fix +1 -1
diff -puN kernel/container_debug.c~lock_fix kernel/container_debug.c
@@ -34,7 +34,7 @@ static u64 taskcount_read(struct contain
 {
 	u64 count;
 	container_lock();
-	count = container_task_count(cont);
+	count = container_task_count(cont, 1);
 	container_unlock();
 	return count;
 }

_

Signature · 2 lines
Regards,
vatsa
#32 lore raw
Quoted 7 lines · Srivatsa Vaddagiri
Quoted 2 lines · Paul Jackson

It builds and boots and mounts the cpuset file system ok. But trying to write the 'mems' file hangs the system hard.

Basically we are attempting a read_lock(&tasklist_lock) in container_task_count() after taking write_lock_irq(&tasklist_lock) in update_nodemask()!

Paul, is there any reason why we need to do a write_lock() on tasklist_lock if we're just trying to block fork, or is it just historical accident? Wouldn't it be fine to do a read_lock()?

Paul

#33 lore raw
Quoted 3 lines · Paul Menage

Paul, is there any reason why we need to do a write_lock() on tasklist_lock if we're just trying to block fork, or is it just historical accident? Wouldn't it be fine to do a read_lock()?

Good point ..read_lock() will probably suffice in update_nodemask which means we don't need the patch I sent earlier.

Paul (Jackson),
This made me see another race in update_nodemask vs fork:

Lets say cpuset CS1 has only one task T1 to begin with.

update_nodemask(CS1) T1 in do_fork()

	CPU0				     CPU1

=============================================================

					cpuset_fork();
					mpol_copy();

ntasks = atomic_read(&cs->count);
[ntasks = 2, accounting new born child T2]
cs->mems_allowed = something;
set_cpuset_being_rebound()

write/read_lock(tasklist_lock);

do_each_thread {

	/* Finds only T1 */

	mmarray[] = ..

} while_each_thread();

write/read_unlock(tasklist_lock);

					write_lock(tasklist_lock);

					/* Add T2, child of T1 to tasklist */

					write_unlock(tasklist_lock);

for (i = 0; i < n; i++) {

	mpol_rebind_mm(..);

}

In this for loop, we migrate only T1's ->mm. T2's->mm isn't migrated AFAICS.

Is that fine?

Signature · 2 lines
Regards,
vatsa
#34 lore raw
In reply to Srivatsa Vaddagiri
Quoted 26 lines · Srivatsa Vaddagiri
Quoted 2 lines · Paul Jackson

It builds and boots and mounts the cpuset file system ok. But trying to write the 'mems' file hangs the system hard.

Basically we are attempting a read_lock(&tasklist_lock) in container_task_count() after taking write_lock_irq(&tasklist_lock) in update_nodemask()!

This patch seems to fix the prb for me:

Fix write_lock() followed by read_lock() bug by introducing a 2nd argument to be passed into container_task_count. Other choice is to introduce a lock and unlocked versions of container_task_count() ..

kernel/container.c renamed from kernel/container.c~lock_fix +1 -2
-int container_task_count(const struct container *cont) {
+int container_task_count(const struct container *cont, int take_lock) {
 	int count = 0;
 	struct task_struct *g, *p;
 	struct container_subsys_state *css;
 	int subsys_id;
 	get_first_subsys(cont, &css, &subsys_id);
 
-	read_lock(&tasklist_lock);

kernel/container.c int container_add_files(struct container · line 1215

Quoted 10 lines
kernel/container.c renamed from kernel/container.c~lock_fix +4 -1
+	if (take_lock)
+		read_lock(&tasklist_lock);
 	do_each_thread(g, p) {
 		if (task_subsys_state(p, subsys_id) == css)
 			count ++;
 	} while_each_thread(g, p);
-	read_unlock(&tasklist_lock);
+	if (take_lock)
+		read_unlock(&tasklist_lock);
 	return count;

Umm, no - please naje two versions with and without the lock. Also Please fix up the codingstyle, the { belongs onto a line of it's own.

#35 lore raw
Quoted 1 line · Christoph Hellwig

Umm, no - please naje two versions with and without the lock.

Ok this patch may not be necessary (given Paul's observation that read_lock will suffice in update_nodemask).

Quoted 1 line

Also Please fix up the codingstyle, the { belongs onto a line of it's own.

Paul, I suspect that there are a number of Codingstyle related changes you may need to address in your next version (including the one pointed by Christoph)

Signature · 2 lines
Regards,
vatsa