Home » Mailing lists » Devel » [PATCH 00/29] Rename Containers to Control Groups
| [PATCH 00/29] Rename Containers to Control Groups [message #20064] |
Tue, 11 September 2007 19:52  |
Paul Menage
Messages: 642
Registered: September 2006
|
Senior Member |
|
|
[ Sending to containers@lists.osdl.org for review before sending on to
linux-kernel and akpm ]
This patchset renames "Task Containers" to "Control Groups", in line with
recent discussions.
These patches are drop-in replacements for those of the same names in
the current -mm tree:
task-containersv11-basic-task-container-framework.patch
task-containersv11-basic-task-container-framework-fix.patch
task-containersv11-add-tasks-file-interface.patch
task-containersv11-add-fork-exit-hooks.patch
task-containersv11-add-container_clone-interface.patch
task-containersv11-add-procfs-interface.patch
task-containersv11-shared-container-subsystem-group-arrays.patch
task-containersv11-shared-container-subsystem-group-arrays-avoid-lockdep-warning.patch
task-containersv11-shared-container-subsystem-group-arrays-include-fix.patch
task-containersv11-automatic-userspace-notification-of-idle-containers.patch
task-containersv11-make-cpusets-a-client-of-containers.patch
task-containersv11-example-cpu-accounting-subsystem.patch
task-containersv11-simple-task-container-debug-info-subsystem.patch
add-containerstats-v3.patch
add-containerstats-v3-fix.patch
containers-implement-namespace-tracking-subsystem.patch
containers-implement-namespace-tracking-subsystem-fix-order-of-container-subsystems-in-init-kconfig.patch
memory-controller-add-documentation.patch
memory-controller-resource-counters-v7.patch
memory-controller-resource-counters-v7-fix.patch
memory-controller-containers-setup-v7.patch
memory-controller-accounting-setup-v7.patch
memory-controller-memory-accounting-v7.patch
memory-controller-task-migration-v7.patch
memory-controller-add-per-container-lru-and-reclaim-v7.patch
memory-controller-add-per-container-lru-and-reclaim-v7-fix.patch
memory-controller-oom-handling-v7.patch
memory-controller-add-switch-to-control-what-type-of-pages-to-limit-v7.patch
memory-controller-make-page_referenced-container-aware-v7.patch
They are constructed by running the following perl script over the
original versions of the patches in 2.6.23-rc4-mm1:
perl -pi -e 's/subcontainer/child cgroup/g; s/(\b|_)container(\b|_(?!of)|fs|s)/$1cgroup$2/g; s/CONTAINER(_|S)/CGROUP$1/g; s/Container/Control Group/g; s/css_group/css_set/g;' patches/*.patch
Owners of other control-group related patches may want to run this
script or some variation of it on their own patches.
--
_______________________________________________
Containers mailing list
Containers@lists.linux-foundation.org
https://lists.linux-foundation.org/mailman/listinfo/containers
|
|
|
|
| [PATCH 14/29] add containerstats v3 [message #20065 is a reply to message #20064] |
Tue, 11 September 2007 19:52  |
Paul Menage
Messages: 642
Registered: September 2006
|
Senior Member |
|
|
From: Balbir Singh <balbir@linux.vnet.ibm.com>
This patch is inspired by the discussion at
http://lkml.org/lkml/2007/4/11/187 and implements per cgroup statistics
as suggested by Andrew Morton in http://lkml.org/lkml/2007/4/11/263. The
patch is on top of 2.6.21-mm1 with Paul's cgroups v9 patches (forward
ported)
This patch implements per cgroup statistics infrastructure and re-uses
code from the taskstats interface. A new set of cgroup operations are
registered with commands and attributes. It should be very easy to
*extend* per cgroup statistics, by adding members to the cgroupstats
structure.
The current model for cgroupstats is a pull, a push model (to post
statistics on interesting events), should be very easy to add. Currently
user space requests for statistics by passing the cgroup file
descriptor. Statistics about the state of all the tasks in the cgroup
is returned to user space.
TODO's/NOTE:
This patch provides an infrastructure for implementing cgroup statistics.
Based on the needs of each controller, we can incrementally add more statistics,
event based support for notification of statistics, accumulation of taskstats
into cgroup statistics in the future.
Sample output
# ./cgroupstats -C /cgroup/a
sleeping 2, blocked 0, running 1, stopped 0, uninterruptible 0
# ./cgroupstats -C /cgroup/
sleeping 154, blocked 0, running 0, stopped 0, uninterruptible 0
If the approach looks good, I'll enhance and post the user space utility for
the same
Feedback, comments, test results are always welcome!
Signed-off-by: Balbir Singh <balbir@linux.vnet.ibm.com>
Cc: Paul Menage <menage@google.com>
Cc: Jay Lan <jlan@engr.sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---
Documentation/accounting/cgroupstats.txt | 27 ++++++
include/linux/Kbuild | 1
include/linux/cgroup.h | 8 ++
include/linux/cgroupstats.h | 70 ++++++++++++++++++
include/linux/delayacct.h | 13 +++
kernel/cgroup.c | 55 ++++++++++++++
kernel/taskstats.c | 66 ++++++++++++++++
7 files changed, 240 insertions(+)
diff -puN /dev/null Documentation/accounting/cgroupstats.txt
--- /dev/null
+++ a/Documentation/accounting/cgroupstats.txt
@@ -0,0 +1,27 @@
+Control Groupstats is inspired by the discussion at
+http://lkml.org/lkml/2007/4/11/187 and implements per cgroup statistics as
+suggested by Andrew Morton in http://lkml.org/lkml/2007/4/11/263.
+
+Per cgroup statistics infrastructure re-uses code from the taskstats
+interface. A new set of cgroup operations are registered with commands
+and attributes specific to cgroups. It should be very easy to
+extend per cgroup statistics, by adding members to the cgroupstats
+structure.
+
+The current model for cgroupstats is a pull, a push model (to post
+statistics on interesting events), should be very easy to add. Currently
+user space requests for statistics by passing the cgroup path.
+Statistics about the state of all the tasks in the cgroup is returned to
+user space.
+
+NOTE: We currently rely on delay accounting for extracting information
+about tasks blocked on I/O. If CONFIG_TASK_DELAY_ACCT is disabled, this
+information will not be available.
+
+To extract cgroup statistics a utility very similar to getdelays.c
+has been developed, the sample output of the utility is shown below
+
+~/balbir/cgroupstats # ./getdelays -C "/cgroup/a"
+sleeping 1, blocked 0, running 1, stopped 0, uninterruptible 0
+~/balbir/cgroupstats # ./getdelays -C "/cgroup"
+sleeping 155, blocked 0, running 1, stopped 0, uninterruptible 2
diff -puN include/linux/Kbuild~add-cgroupstats-v3 include/linux/Kbuild
--- a/include/linux/Kbuild~add-cgroupstats-v3
+++ a/include/linux/Kbuild
@@ -47,6 +47,7 @@ header-y += coda_psdev.h
header-y += coff.h
header-y += comstats.h
header-y += const.h
+header-y += cgroupstats.h
header-y += cycx_cfm.h
header-y += dlm_device.h
header-y += dlm_netlink.h
diff -puN include/linux/cgroup.h~add-cgroupstats-v3 include/linux/cgroup.h
--- a/include/linux/cgroup.h~add-cgroupstats-v3
+++ a/include/linux/cgroup.h
@@ -13,6 +13,7 @@
#include <linux/cpumask.h>
#include <linux/nodemask.h>
#include <linux/rcupdate.h>
+#include <linux/cgroupstats.h>
#ifdef CONFIG_CGROUPS
@@ -29,6 +30,8 @@ extern void cgroup_fork(struct task_s
extern void cgroup_fork_callbacks(struct task_struct *p);
extern void cgroup_post_fork(struct task_struct *p);
extern void cgroup_exit(struct task_struct *p, int run_callbacks);
+extern int cgroupstats_build(struct cgroupstats *stats,
+ struct dentry *dentry);
extern struct file_operations proc_cgroup_operations;
@@ -305,6 +308,11 @@ static inline void cgroup_exit(struct
static inline void cgroup_lock(void) {}
static inline void cgroup_unlock(void) {}
+static inline int cgroupstats_build(struct cgroupstats *stats,
+ struct dentry *dentry)
+{
+ return -EINVAL;
+}
#endif /* !CONFIG_CGROUPS */
diff -puN /dev/null include/linux/cgroupstats.h
--- /dev/null
+++ a/include/linux/cgroupstats.h
@@ -0,0 +1,70 @@
+/* cgroupstats.h - exporting per-cgroup statistics
+ *
+ * Copyright IBM Corporation, 2007
+ * Author Balbir Singh <balbir@linux.vnet.ibm.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of version 2.1 of the GNU Lesser General Public License
+ * as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it would be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
+ */
+
+#ifndef _LINUX_CGROUPSTATS_H
+#define _LINUX_CGROUPSTATS_H
+
+#include <linux/taskstats.h>
+
+/*
+ * Data shared between user space and kernel space on a per cgroup
+ * basis. This data is shared using taskstats.
+ *
+ * Most of these states are derived by looking at the task->state value
+ * For the nr_io_wait state, a flag in the delay accounting structure
+ * indicates that the task is waiting on IO
+ *
+ * Each member is aligned to a 8 byte boundary.
+ */
+struct cgroupstats {
+ __u64 nr_sleeping; /* Number of tasks sleeping */
+ __u64 nr_running; /* Number of tasks running */
+ __u64 nr_stopped; /* Number of tasks in stopped state */
+ __u64 nr_uninterruptible; /* Number of tasks in uninterruptible */
+ /* state */
+ __u64 nr_io_wait; /* Number of tasks waiting on IO */
+};
+
+/*
+ * Commands sent from userspace
+ * Not versioned. New commands should only be inserted at the enum's end
+ * prior to __CGROUPSTATS_CMD_MAX
+ */
+
+enum {
+ CGROUPSTATS_CMD_UNSPEC = __TASKSTATS_CMD_MAX, /* Reserved */
+ CGROUPSTATS_CMD_GET, /* user->kernel request/get-response */
+ CGROUPSTATS_CMD_NEW, /* kernel->user event */
+ __CGROUPSTATS_CMD_MAX,
+};
+
+#define CGROUPSTATS_CMD_MAX (__CGROUPSTATS_CMD_MAX - 1)
+
+enum {
+ CGROUPSTATS_TYPE_UNSPEC = 0, /* Reserved */
+ CGROUPSTATS_TYPE_CGROUP_STATS, /* contains name + stats */
+ __CGROUPSTATS_TYPE_MAX,
+};
+
+#define CGROUPSTATS_TYPE_MAX (__CGROUPSTATS_TYPE_MAX - 1)
+
+enum {
+ CGROUPSTATS_CMD_ATTR_UNSPEC = 0,
+ CGROUPSTATS_CMD_ATTR_FD,
+ __CGROUPSTATS_CMD_ATTR_MAX,
+};
+
+#define CGROUPSTATS_CMD_ATTR_MAX (__CGROUPSTATS_CMD_ATTR_MAX - 1)
+
+#endif /* _LINUX_CGROUPSTATS_H */
diff -puN include/linux/delayacct.h~add-cgroupstats-v3 include/linux/delayacct.h
--- a/include/linux/delayacct.h~add-cgroupstats-v3
+++ a/include/linux/delayacct.h
@@ -26,6 +26,7 @@
* Used to set current->delays->flags
*/
#define DELAYACCT_PF_SWAPIN 0x00000001 /* I am doing a swapin */
+#define DELAYACCT_PF_BLKIO 0x00000002 /* I am waiting on IO */
#ifdef CONFIG_TASK_DELAY_ACCT
@@ -39,6 +40,14 @@ extern void __delayacct_blkio_end(void);
extern int __delayacct_add_tsk(struct taskstats *, struct task_struct *);
extern __u64 __delayacct_blkio_ticks(struct task_struct *);
+static inline int delayacct_is_task_waiting_on_io(struct task_struct *p)
+{
+ if (p->delays)
+ return (p->delays->flags & DELAYACCT_PF_BLKIO);
+ else
+ return 0;
+}
+
static inline void delayacct_set_flag(int flag)
{
if (current->delays)
@@ -71,6 +80,7 @@ static inline void delayacct_tsk_free(st
static inline void delayacct_blkio_start(void)
{
+ delayacct_set_flag(DELAYACCT_PF_BLKIO);
if (current->delays)
__delayacct_blkio_start();
}
@@ -79,6 +89,7 @@ static inline void delayacct_blkio_end(v
{
if (current->delays)
__delayacct_blkio_end();
+ delayacct_clear_flag(DELAYACCT_PF_BLKIO);
}
static inline int delayacct_add_tsk(struct taskstats *d,
@@ -116,6 +127,8 @@ static inline int delayacct_add_tsk(stru
{ return 0; }
static inline __u64 delayacct_blkio_ticks(struct task_struct *tsk)
{ return 0; }
+static inline int delayacct_is_task_waiting_on_io(struct task_struct *p)
+{ return 0; }
#endif /* CONFIG_TASK_DELAY_ACCT */
#endif
diff -puN kernel/cgroup.c~add-cgroupstats-v3 kernel/cgroup.c
--- a/kernel/cgroup.c~add-cgroupstats-v3
+++ a/kernel/cgroup.c
@@ -42,6 +42,9 @@
#include <linux/spinlock.h>
#include <linux/string.h>
#include <linux/sort.h>
+#include <linux/delayacct.h>
+#include <linux/cgroupstats.h>
+
#include <asm/atomic.h>
static DEFINE_MUTEX(cgroup_mutex);
@@ -1739,6 +1742,58 @@ static int pid_array_load(pid_t *pidarra
return n;
}
+/**
+ * Build and fill cgroupstats so that taskstats can export it to user
+ * space.
+ *
+ * @stats: cgroupstats to fill information into
+ * @dentry: A dentry entry belonging to the cgroup for which stats have
+ * been requested.
+ */
+int cgroupstats_build(struct cgroupstats *stats, struct dentry *dentry)
+{
+ int ret = -EINVAL;
+ struct cgroup *cont;
+ struct cgroup_iter it;
+ struct task_struct *tsk;
+ /*
+ * Validate dentry by checking the superblock operations
+ */
+ if (dentry->d_sb->s_op != &cgroup_ops)
+ g
...
|
|
|
|
|
|
| [PATCH 16/29] containers implement namespace tracking subsystem [message #20067 is a reply to message #20064] |
Tue, 11 September 2007 19:52 |
Paul Menage
Messages: 642
Registered: September 2006
|
Senior Member |
|
|
From: "Serge E. Hallyn" <serue@us.ibm.com>
When a task enters a new namespace via a clone() or unshare(), a new cgroup
is created and the task moves into it.
This version names cgroups which are automatically created using
cgroup_clone() as "node_<pid>" where pid is the pid of the unsharing or
cloned process. (Thanks Pavel for the idea) This is safe because if the
process unshares again, it will create
/cgroups/(...)/node_<pid>/node_<pid>
The only possibilities (AFAICT) for a -EEXIST on unshare are
1. pid wraparound
2. a process fails an unshare, then tries again.
Case 1 is unlikely enough that I ignore it (at least for now). In case 2, the
node_<pid> will be empty and can be rmdir'ed to make the subsequent unshare()
succeed.
Changelog:
Name cloned cgroups as "node_<pid>".
Signed-off-by: Serge E. Hallyn <serue@us.ibm.com>
Cc: Paul Menage <menage@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---
include/linux/cgroup_subsys.h | 6 +
include/linux/nsproxy.h | 7 ++
init/Kconfig | 9 ++
kernel/Makefile | 1
kernel/ns_cgroup.c | 100 +++++++++++++++++++++++++++++
kernel/nsproxy.c | 17 ++++
6 files changed, 139 insertions(+), 1 deletion(-)
diff -puN include/linux/cgroup_subsys.h~cgroups-implement-namespace-tracking-subsystem include/linux/cgroup_subsys.h
--- a/include/linux/cgroup_subsys.h~cgroups-implement-namespace-tracking-subsystem
+++ a/include/linux/cgroup_subsys.h
@@ -24,3 +24,9 @@ SUBSYS(debug)
#endif
/* */
+
+#ifdef CONFIG_CGROUP_NS
+SUBSYS(ns)
+#endif
+
+/* */
diff -puN include/linux/nsproxy.h~cgroups-implement-namespace-tracking-subsystem include/linux/nsproxy.h
--- a/include/linux/nsproxy.h~cgroups-implement-namespace-tracking-subsystem
+++ a/include/linux/nsproxy.h
@@ -55,4 +55,11 @@ static inline void exit_task_namespaces(
put_nsproxy(ns);
}
}
+
+#ifdef CONFIG_CGROUP_NS
+int ns_cgroup_clone(struct task_struct *tsk);
+#else
+static inline int ns_cgroup_clone(struct task_struct *tsk) { return 0; }
+#endif
+
#endif
diff -puN init/Kconfig~cgroups-implement-namespace-tracking-subsystem init/Kconfig
--- a/init/Kconfig~cgroups-implement-namespace-tracking-subsystem
+++ a/init/Kconfig
@@ -323,6 +323,15 @@ 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 CGROUP_NS
+ bool "Namespace cgroup subsystem"
+ select CGROUPS
+ help
+ Provides a simple namespace cgroup subsystem to
+ provide hierarchical naming of sets of namespaces,
+ for instance virtual servers and checkpoint/restart
+ jobs.
+
config PROC_PID_CPUSET
bool "Include legacy /proc/<pid>/cpuset file"
depends on CPUSETS
diff -puN kernel/Makefile~cgroups-implement-namespace-tracking-subsystem kernel/Makefile
--- a/kernel/Makefile~cgroups-implement-namespace-tracking-subsystem
+++ a/kernel/Makefile
@@ -42,6 +42,7 @@ obj-$(CONFIG_CGROUPS) += cgroup.o
obj-$(CONFIG_CGROUP_DEBUG) += cgroup_debug.o
obj-$(CONFIG_CPUSETS) += cpuset.o
obj-$(CONFIG_CGROUP_CPUACCT) += cpu_acct.o
+obj-$(CONFIG_CGROUP_NS) += ns_cgroup.o
obj-$(CONFIG_IKCONFIG) += configs.o
obj-$(CONFIG_STOP_MACHINE) += stop_machine.o
obj-$(CONFIG_AUDIT) += audit.o auditfilter.o
diff -puN /dev/null kernel/ns_cgroup.c
--- /dev/null
+++ a/kernel/ns_cgroup.c
@@ -0,0 +1,100 @@
+/*
+ * ns_cgroup.c - namespace cgroup subsystem
+ *
+ * Copyright 2006, 2007 IBM Corp
+ */
+
+#include <linux/module.h>
+#include <linux/cgroup.h>
+#include <linux/fs.h>
+
+struct ns_cgroup {
+ struct cgroup_subsys_state css;
+ spinlock_t lock;
+};
+
+struct cgroup_subsys ns_subsys;
+
+static inline struct ns_cgroup *cgroup_to_ns(
+ struct cgroup *cgroup)
+{
+ return container_of(cgroup_subsys_state(cgroup, ns_subsys_id),
+ struct ns_cgroup, css);
+}
+
+int ns_cgroup_clone(struct task_struct *task)
+{
+ return cgroup_clone(task, &ns_subsys);
+}
+
+/*
+ * Rules:
+ * 1. you can only enter a cgroup which is a child of your current
+ * cgroup
+ * 2. you can only place another process into a cgroup if
+ * a. you have CAP_SYS_ADMIN
+ * b. your cgroup is an ancestor of task's destination cgroup
+ * (hence either you are in the same cgroup as task, or in an
+ * ancestor cgroup thereof)
+ */
+static int ns_can_attach(struct cgroup_subsys *ss,
+ struct cgroup *new_cgroup, struct task_struct *task)
+{
+ struct cgroup *orig;
+
+ if (current != task) {
+ if (!capable(CAP_SYS_ADMIN))
+ return -EPERM;
+
+ if (!cgroup_is_descendant(new_cgroup))
+ return -EPERM;
+ }
+
+ if (atomic_read(&new_cgroup->count) != 0)
+ return -EPERM;
+
+ orig = task_cgroup(task, ns_subsys_id);
+ if (orig && orig != new_cgroup->parent)
+ return -EPERM;
+
+ return 0;
+}
+
+/*
+ * Rules: you can only create a cgroup if
+ * 1. you are capable(CAP_SYS_ADMIN)
+ * 2. the target cgroup is a descendant of your own cgroup
+ */
+static struct cgroup_subsys_state *ns_create(struct cgroup_subsys *ss,
+ struct cgroup *cgroup)
+{
+ struct ns_cgroup *ns_cgroup;
+
+ if (!capable(CAP_SYS_ADMIN))
+ return ERR_PTR(-EPERM);
+ if (!cgroup_is_descendant(cgroup))
+ return ERR_PTR(-EPERM);
+
+ ns_cgroup = kzalloc(sizeof(*ns_cgroup), GFP_KERNEL);
+ if (!ns_cgroup)
+ return ERR_PTR(-ENOMEM);
+ spin_lock_init(&ns_cgroup->lock);
+ return &ns_cgroup->css;
+}
+
+static void ns_destroy(struct cgroup_subsys *ss,
+ struct cgroup *cgroup)
+{
+ struct ns_cgroup *ns_cgroup;
+
+ ns_cgroup = cgroup_to_ns(cgroup);
+ kfree(ns_cgroup);
+}
+
+struct cgroup_subsys ns_subsys = {
+ .name = "ns",
+ .can_attach = ns_can_attach,
+ .create = ns_create,
+ .destroy = ns_destroy,
+ .subsys_id = ns_subsys_id,
+};
diff -puN kernel/nsproxy.c~cgroups-implement-namespace-tracking-subsystem kernel/nsproxy.c
--- a/kernel/nsproxy.c~cgroups-implement-namespace-tracking-subsystem
+++ a/kernel/nsproxy.c
@@ -146,7 +146,14 @@ int copy_namespaces(unsigned long flags,
goto out;
}
+ err = ns_cgroup_clone(tsk);
+ if (err) {
+ put_nsproxy(new_ns);
+ goto out;
+ }
+
tsk->nsproxy = new_ns;
+
out:
put_nsproxy(old_ns);
return err;
@@ -185,8 +192,16 @@ int unshare_nsproxy_namespaces(unsigned
*new_nsp = create_new_namespaces(unshare_flags, current,
new_fs ? new_fs : current->fs);
- if (IS_ERR(*new_nsp))
+ if (IS_ERR(*new_nsp)) {
err = PTR_ERR(*new_nsp);
+ goto out;
+ }
+
+ err = ns_cgroup_clone(current);
+ if (err)
+ put_nsproxy(*new_nsp);
+
+out:
return err;
}
_
--
_______________________________________________
Containers mailing list
Containers@lists.linux-foundation.org
https://lists.linux-foundation.org/mailman/listinfo/containers
|
|
|
|
| [PATCH 24/29] memory controller task migration v7 [message #20068 is a reply to message #20064] |
Tue, 11 September 2007 19:53 |
Paul Menage
Messages: 642
Registered: September 2006
|
Senior Member |
|
|
From: Balbir Singh <balbir@linux.vnet.ibm.com>
Allow tasks to migrate from one cgroup to the other. We migrate
mm_struct's mem_cgroup only when the thread group id migrates.
Signed-off-by: Balbir Singh <balbir@linux.vnet.ibm.com>
Cc: Pavel Emelianov <xemul@openvz.org>
Cc: Paul Menage <menage@google.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Cc: Kirill Korotaev <dev@sw.ru>
Cc: Herbert Poetzl <herbert@13thfloor.at>
Cc: David Rientjes <rientjes@google.com>
Cc: Vaidyanathan Srinivasan <svaidy@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---
mm/memcontrol.c | 35 +++++++++++++++++++++++++++++++++++
1 files changed, 35 insertions(+)
diff -puN mm/memcontrol.c~memory-controller-task-migration-v7 mm/memcontrol.c
--- a/mm/memcontrol.c~memory-controller-task-migration-v7
+++ a/mm/memcontrol.c
@@ -326,11 +326,46 @@ static int mem_cgroup_populate(struct
ARRAY_SIZE(mem_cgroup_files));
}
+static void mem_cgroup_move_task(struct cgroup_subsys *ss,
+ struct cgroup *cont,
+ struct cgroup *old_cont,
+ struct task_struct *p)
+{
+ struct mm_struct *mm;
+ struct mem_cgroup *mem, *old_mem;
+
+ mm = get_task_mm(p);
+ if (mm == NULL)
+ return;
+
+ mem = mem_cgroup_from_cont(cont);
+ old_mem = mem_cgroup_from_cont(old_cont);
+
+ if (mem == old_mem)
+ goto out;
+
+ /*
+ * Only thread group leaders are allowed to migrate, the mm_struct is
+ * in effect owned by the leader
+ */
+ if (p->tgid != p->pid)
+ goto out;
+
+ css_get(&mem->css);
+ rcu_assign_pointer(mm->mem_cgroup, mem);
+ css_put(&old_mem->css);
+
+out:
+ mmput(mm);
+ return;
+}
+
struct cgroup_subsys mem_cgroup_subsys = {
.name = "memory",
.subsys_id = mem_cgroup_subsys_id,
.create = mem_cgroup_create,
.destroy = mem_cgroup_destroy,
.populate = mem_cgroup_populate,
+ .attach = mem_cgroup_move_task,
.early_init = 1,
};
_
--
_______________________________________________
Containers mailing list
Containers@lists.linux-foundation.org
https://lists.linux-foundation.org/mailman/listinfo/containers
|
|
|
|
|
|
| [PATCH 08/29] task containersv11 shared container subsystem group arrays avoid lockdep warning [message #20070 is a reply to message #20064] |
Tue, 11 September 2007 19:52 |
Paul Menage
Messages: 642
Registered: September 2006
|
Senior Member |
|
|
From: Paul Menage <menage@google.com>
I think this is the right way to handle the lockdep false-positive in the
current cgroups patches, but I'm not that familiar with lockdep so any
suggestions for a better approach are welcomed.
In order to avoid a false-positive lockdep warning, we lock the root inode
of a new filesystem mount prior to taking cgroup_mutex, to preserve the
invariant that cgroup_mutex nests inside inode->i_mutex. In order to
prevent a lockdep false positive when locking i_mutex on a newly-created
cgroup directory inode we use mutex_lock_nested(), with a nesting level
of I_MUTEX_CHILD since the new inode will ultimately be a child directory
of the parent whose i_mutex is nested outside of cgroup_mutex.
Signed-off-by: Paul Menage <menage@google.com>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---
kernel/cgroup.c | 17 +++++++----------
1 files changed, 7 insertions(+), 10 deletions(-)
diff -puN kernel/cgroup.c~task-cgroupsv11-shared-cgroup-subsystem-group-arrays-avoid-lockdep-warning kernel/cgroup.c
--- a/kernel/cgroup.c~task-cgroupsv11-shared-cgroup-subsystem-group-arrays-avoid-lockdep-warning
+++ a/kernel/cgroup.c
@@ -867,13 +867,16 @@ static int cgroup_get_sb(struct file_
} else {
/* New superblock */
struct cgroup *cont = &root->top_cgroup;
+ struct inode *inode;
BUG_ON(sb->s_root != NULL);
ret = cgroup_get_rootdir(sb);
if (ret)
goto drop_new_super;
+ inode = sb->s_root->d_inode;
+ mutex_lock(&inode->i_mutex);
mutex_lock(&cgroup_mutex);
/*
@@ -886,12 +889,14 @@ static int cgroup_get_sb(struct file_
ret = allocate_cg_links(css_set_count, &tmp_cg_links);
if (ret) {
mutex_unlock(&cgroup_mutex);
+ mutex_unlock(&inode->i_mutex);
goto drop_new_super;
}
ret = rebind_subsystems(root, root->subsys_bits);
if (ret == -EBUSY) {
mutex_unlock(&cgroup_mutex);
+ mutex_unlock(&inode->i_mutex);
goto drop_new_super;
}
@@ -931,16 +936,8 @@ static int cgroup_get_sb(struct file_
BUG_ON(!list_empty(&cont->children));
BUG_ON(root->number_of_cgroups != 1);
- /*
- * I believe that it's safe to nest i_mutex inside
- * cgroup_mutex in this case, since no-one else can
- * be accessing this directory yet. But we still need
- * to teach lockdep that this is the case - currently
- * a cgroupfs remount triggers a lockdep warning
- */
- mutex_lock(&cont->dentry->d_inode->i_mutex);
cgroup_populate_dir(cont);
- mutex_unlock(&cont->dentry->d_inode->i_mutex);
+ mutex_unlock(&inode->i_mutex);
mutex_unlock(&cgroup_mutex);
}
@@ -1358,7 +1355,7 @@ static int cgroup_create_file(struct
/* start with the directory inode held, so that we can
* populate it without racing with another mkdir */
- mutex_lock(&inode->i_mutex);
+ mutex_lock_nested(&inode->i_mutex, I_MUTEX_CHILD);
} else if (S_ISREG(mode)) {
inode->i_size = 0;
inode->i_fop = &cgroup_file_operations;
_
--
_______________________________________________
Containers mailing list
Containers@lists.linux-foundation.org
https://lists.linux-foundation.org/mailman/listinfo/containers
|
|
|
|
|
|
| [PATCH 18/29] memory controller add documentation [message #20072 is a reply to message #20064] |
Tue, 11 September 2007 19:52 |
Paul Menage
Messages: 642
Registered: September 2006
|
Senior Member |
|
|
From: Balbir Singh <balbir@linux.vnet.ibm.com>
Signed-off-by: Balbir Singh <balbir@linux.vnet.ibm.com>
Cc: Pavel Emelianov <xemul@openvz.org>
Cc: Paul Menage <menage@google.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Cc: Kirill Korotaev <dev@sw.ru>
Cc: Herbert Poetzl <herbert@13thfloor.at>
Cc: David Rientjes <rientjes@google.com>
Cc: Vaidyanathan Srinivasan <svaidy@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---
Documentation/controllers/memory.txt | 259 +++++++++++++++++++++++++
1 files changed, 259 insertions(+)
diff -puN /dev/null Documentation/controllers/memory.txt
--- /dev/null
+++ a/Documentation/controllers/memory.txt
@@ -0,0 +1,259 @@
+Memory Controller
+
+Salient features
+
+a. Enable control of both RSS (mapped) and Page Cache (unmapped) pages
+b. The infrastructure allows easy addition of other types of memory to control
+c. Provides *zero overhead* for non memory controller users
+d. Provides a double LRU: global memory pressure causes reclaim from the
+ global LRU; a cgroup on hitting a limit, reclaims from the per
+ cgroup LRU
+
+NOTE: Page Cache (unmapped) also includes Swap Cache pages as a subset
+and will not be referred to explicitly in the rest of the documentation.
+
+Benefits and Purpose of the memory controller
+
+The memory controller isolates the memory behaviour of a group of tasks
+from the rest of the system. The article on LWN [12] mentions some probable
+uses of the memory controller. The memory controller can be used to
+
+a. Isolate an application or a group of applications
+ Memory hungry applications can be isolated and limited to a smaller
+ amount of memory.
+b. Create a cgroup with limited amount of memory, this can be used
+ as a good alternative to booting with mem=XXXX.
+c. Virtualization solutions can control the amount of memory they want
+ to assign to a virtual machine instance.
+d. A CD/DVD burner could control the amount of memory used by the
+ rest of the system to ensure that burning does not fail due to lack
+ of available memory.
+e. There are several other use cases, find one or use the controller just
+ for fun (to learn and hack on the VM subsystem).
+
+1. History
+
+The memory controller has a long history. A request for comments for the memory
+controller was posted by Balbir Singh [1]. At the time the RFC was posted
+there were several implementations for memory control. The goal of the
+RFC was to build consensus and agreement for the minimal features required
+for memory control. The first RSS controller was posted by Balbir Singh[2]
+in Feb 2007. Pavel Emelianov [3][4][5] has since posted three versions of the
+RSS controller. At OLS, at the resource management BoF, everyone suggested
+that we handle both page cache and RSS together. Another request was raised
+to allow user space handling of OOM. The current memory controller is
+at version 6; it combines both mapped (RSS) and unmapped Page
+Cache Control [11].
+
+2. Memory Control
+
+Memory is a unique resource in the sense that it is present in a limited
+amount. If a task requires a lot of CPU processing, the task can spread
+its processing over a period of hours, days, months or years, but with
+memory, the same physical memory needs to be reused to accomplish the task.
+
+The memory controller implementation has been divided into phases. These
+are:
+
+1. Memory controller
+2. mlock(2) controller
+3. Kernel user memory accounting and slab control
+4. user mappings length controller
+
+The memory controller is the first controller developed.
+
+2.1. Design
+
+The core of the design is a counter called the res_counter. The res_counter
+tracks the current memory usage and limit of the group of processes associated
+with the controller. Each cgroup has a memory controller specific data
+structure (mem_cgroup) associated with it.
+
+2.2. Accounting
+
+ +--------------------+
+ | mem_cgroup |
+ | (res_counter) |
+ +--------------------+
+ / ^ \
+ / | \
+ +---------------+ | +---------------+
+ | mm_struct | |.... | mm_struct |
+ | | | | |
+ +---------------+ | +---------------+
+ |
+ + --------------+
+ |
+ +---------------+ +------+--------+
+ | page +----------> page_cgroup|
+ | | | |
+ +---------------+ +---------------+
+
+ (Figure 1: Hierarchy of Accounting)
+
+
+Figure 1 shows the important aspects of the controller
+
+1. Accounting happens per cgroup
+2. Each mm_struct knows about which cgroup it belongs to
+3. Each page has a pointer to the page_cgroup, which in turn knows the
+ cgroup it belongs to
+
+The accounting is done as follows: mem_cgroup_charge() is invoked to setup
+the necessary data structures and check if the cgroup that is being charged
+is over its limit. If it is then reclaim is invoked on the cgroup.
+More details can be found in the reclaim section of this document.
+If everything goes well, a page meta-data-structure called page_cgroup is
+allocated and associated with the page. This routine also adds the page to
+the per cgroup LRU.
+
+2.2.1 Accounting details
+
+All mapped pages (RSS) and unmapped user pages (Page Cache) are accounted.
+RSS pages are accounted at the time of page_add_*_rmap() unless they've already
+been accounted for earlier. A file page will be accounted for as Page Cache;
+it's mapped into the page tables of a process, duplicate accounting is carefully
+avoided. Page Cache pages are accounted at the time of add_to_page_cache().
+The corresponding routines that remove a page from the page tables or removes
+a page from Page Cache is used to decrement the accounting counters of the
+cgroup.
+
+2.3 Shared Page Accounting
+
+Shared pages are accounted on the basis of the first touch approach. The
+cgroup that first touches a page is accounted for the page. The principle
+behind this approach is that a cgroup that aggressively uses a shared
+page will eventually get charged for it (once it is uncharged from
+the cgroup that brought it in -- this will happen on memory pressure).
+
+2.4 Reclaim
+
+Each cgroup maintains a per cgroup LRU that consists of an active
+and inactive list. When a cgroup goes over its limit, we first try
+to reclaim memory from the cgroup so as to make space for the new
+pages that the cgroup has touched. If the reclaim is unsuccessful,
+an OOM routine is invoked to select and kill the bulkiest task in the
+cgroup.
+
+The reclaim algorithm has not been modified for cgroups, except that
+pages that are selected for reclaiming come from the per cgroup LRU
+list.
+
+2. Locking
+
+The memory controller uses the following hierarchy
+
+1. zone->lru_lock is used for selecting pages to be isolated
+2. mem->lru_lock protects the per cgroup LRU
+3. lock_page_cgroup() is used to protect page->page_cgroup
+
+3. User Interface
+
+0. Configuration
+
+a. Enable CONFIG_CGROUPS
+b. Enable CONFIG_RESOURCE_COUNTERS
+c. Enable CONFIG_CGROUP_MEM_CONT
+
+1. Prepare the cgroups
+# mkdir -p /cgroups
+# mount -t cgroup none /cgroups -o memory
+
+2. Make the new group and move bash into it
+# mkdir /cgroups/0
+# echo $$ > /cgroups/0/tasks
+
+Since now we're in the 0 cgroup,
+We can alter the memory limit:
+# echo -n 6000 > /cgroups/0/memory.limit
+
+We can check the usage:
+# cat /cgroups/0/memory.usage
+25
+
+The memory.failcnt field gives the number of times that the cgroup limit was
+exceeded.
+
+4. Testing
+
+Balbir posted lmbench, AIM9, LTP and vmmstress results [10] and [11].
+Apart from that v6 has been tested with several applications and regular
+daily use. The controller has also been tested on the PPC64, x86_64 and
+UML platforms.
+
+4.1 Troubleshooting
+
+Sometimes a user might find that the application under a cgroup is
+terminated. There are several causes for this:
+
+1. The cgroup limit is too low (just too low to do anything useful)
+2. The user is using anonymous memory and swap is turned off or too low
+
+A sync followed by echo 1 > /proc/sys/vm/drop_caches will help get rid of
+some of the pages cached in the cgroup (page cache pages).
+
+4.2 Task migration
+
+When a task migrates from one cgroup to another, it's charge is not
+carried forward. The pages allocated from the original cgroup still
+remain charged to it, the charge is dropped when the page is freed or
+reclaimed.
+
+4.3 Removing a cgroup
+
+A cgroup can be removed by rmdir, but as discussed in sections 4.1 and 4.2, a
+cgroup might have some charge associated with it, even though all
+tasks have migrated away from it. If some pages are still left, after following
+the steps listed in sections 4.1 and 4.2, check the Swap Cache usage in
+/proc/meminfo to see if the Swap Cache usage is showing up in the
+cgroups memory.usage counter. A simple test of swapoff -a and swapon -a
+should free any pending Swap Cache usage.
+
+4.4 Choosing what to account -- Page Cache (unmapped) vs RSS (mapped)?
+
+The type of memory accounted by the cgroup can be limited to just
+mapped pages by writing "1" to memory.control_type field
+
+echo -n 1 > memory.control_type
+
+5. TODO
+
+1. Add support for accounting huge pages (as a separate controller)
+2. Improve the user interface to accept/display memory limits in KB or MB
+ rather than pages (since page sizes can differ across platforms/machines).
+3. Make cgroup lists per-zone
+4. Make per-cgroup scanner reclaim not-shared pages first
+5. Teach controller to account for shared-pages
+6. Start reclamation when
...
|
|
|
|
| [PATCH 03/29] task containersv11 add tasks file interface [message #20073 is a reply to message #20064] |
Tue, 11 September 2007 19:52 |
Paul Menage
Messages: 642
Registered: September 2006
|
Senior Member |
|
|
From: Paul Menage <menage@google.com>
Add the per-directory "tasks" file for cgroupfs mounts; this allows the
user to determine which tasks are members of a cgroup by reading a
cgroup's "tasks", and to move a task into a cgroup by writing its pid to
its "tasks".
Signed-off-by: Paul Menage <menage@google.com>
Cc: Serge E. Hallyn <serue@us.ibm.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Dave Hansen <haveblue@us.ibm.com>
Cc: Balbir Singh <balbir@in.ibm.com>
Cc: Paul Jackson <pj@sgi.com>
Cc: Kirill Korotaev <dev@openvz.org>
Cc: Herbert Poetzl <herbert@13thfloor.at>
Cc: Srivatsa Vaddagiri <vatsa@in.ibm.com>
Cc: Cedric Le Goater <clg@fr.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---
include/linux/cgroup.h | 10 +
kernel/cgroup.c | 359 +++++++++++++++++++++++++++++++++++-
2 files changed, 368 insertions(+), 1 deletion(-)
diff -puN include/linux/cgroup.h~task-cgroupsv11-add-tasks-file-interface include/linux/cgroup.h
--- a/include/linux/cgroup.h~task-cgroupsv11-add-tasks-file-interface
+++ a/include/linux/cgroup.h
@@ -144,6 +144,16 @@ int cgroup_is_removed(const struct co
int cgroup_path(const struct cgroup *cont, char *buf, int buflen);
+int __cgroup_task_count(const struct cgroup *cont);
+static inline int cgroup_task_count(const struct cgroup *cont)
+{
+ int task_count;
+ rcu_read_lock();
+ task_count = __cgroup_task_count(cont);
+ rcu_read_unlock();
+ return task_count;
+}
+
/* Return true if the cgroup is a descendant of the current cgroup */
int cgroup_is_descendant(const struct cgroup *cont);
diff -puN kernel/cgroup.c~task-cgroupsv11-add-tasks-file-interface kernel/cgroup.c
--- a/kernel/cgroup.c~task-cgroupsv11-add-tasks-file-interface
+++ a/kernel/cgroup.c
@@ -40,7 +40,7 @@
#include <linux/magic.h>
#include <linux/spinlock.h>
#include <linux/string.h>
-
+#include <linux/sort.h>
#include <asm/atomic.h>
/* Generate an array of cgroup subsystem pointers */
@@ -713,6 +713,127 @@ int cgroup_path(const struct containe
return 0;
}
+/*
+ * Return the first subsystem attached to a cgroup's hierarchy, and
+ * its subsystem id.
+ */
+
+static void get_first_subsys(const struct cgroup *cont,
+ struct cgroup_subsys_state **css, int *subsys_id)
+{
+ const struct cgroupfs_root *root = cont->root;
+ const struct cgroup_subsys *test_ss;
+ BUG_ON(list_empty(&root->subsys_list));
+ test_ss = list_entry(root->subsys_list.next,
+ struct cgroup_subsys, sibling);
+ if (css) {
+ *css = cont->subsys[test_ss->subsys_id];
+ BUG_ON(!*css);
+ }
+ if (subsys_id)
+ *subsys_id = test_ss->subsys_id;
+}
+
+/*
+ * Attach task 'tsk' to cgroup 'cont'
+ *
+ * Call holding cgroup_mutex. May take task_lock of
+ * the task 'pid' during call.
+ */
+static int attach_task(struct cgroup *cont, struct task_struct *tsk)
+{
+ int retval = 0;
+ struct cgroup_subsys *ss;
+ struct cgroup *oldcont;
+ struct css_set *cg = &tsk->cgroups;
+ struct cgroupfs_root *root = cont->root;
+ int i;
+ int subsys_id;
+
+ get_first_subsys(cont, NULL, &subsys_id);
+
+ /* Nothing to do if the task is already in that cgroup */
+ oldcont = task_cgroup(tsk, subsys_id);
+ if (cont == oldcont)
+ return 0;
+
+ for_each_subsys(root, ss) {
+ if (ss->can_attach) {
+ retval = ss->can_attach(ss, cont, tsk);
+ if (retval) {
+ return retval;
+ }
+ }
+ }
+
+ task_lock(tsk);
+ if (tsk->flags & PF_EXITING) {
+ task_unlock(tsk);
+ return -ESRCH;
+ }
+ /* Update the css_set pointers for the subsystems in this
+ * hierarchy */
+ for (i = 0; i < CGROUP_SUBSYS_COUNT; i++) {
+ if (root->subsys_bits & (1ull << i)) {
+ /* Subsystem is in this hierarchy. So we want
+ * the subsystem state from the new
+ * cgroup. Transfer the refcount from the
+ * old to the new */
+ atomic_inc(&cont->count);
+ atomic_dec(&cg->subsys[i]->cgroup->count);
+ rcu_assign_pointer(cg->subsys[i], cont->subsys[i]);
+ }
+ }
+ task_unlock(tsk);
+
+ for_each_subsys(root, ss) {
+ if (ss->attach) {
+ ss->attach(ss, cont, oldcont, tsk);
+ }
+ }
+
+ synchronize_rcu();
+ return 0;
+}
+
+/*
+ * Attach task with pid 'pid' to cgroup 'cont'. Call with
+ * cgroup_mutex, may take task_lock of task
+ */
+static int attach_task_by_pid(struct cgroup *cont, char *pidbuf)
+{
+ pid_t pid;
+ struct task_struct *tsk;
+ int ret;
+
+ if (sscanf(pidbuf, "%d", &pid) != 1)
+ return -EIO;
+
+ if (pid) {
+ rcu_read_lock();
+ tsk = find_task_by_pid(pid);
+ if (!tsk || tsk->flags & PF_EXITING) {
+ rcu_read_unlock();
+ return -ESRCH;
+ }
+ get_task_struct(tsk);
+ rcu_read_unlock();
+
+ 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 cgroup file system */
enum cgroup_filetype {
@@ -721,6 +842,55 @@ enum cgroup_filetype {
FILE_TASKLIST,
};
+static ssize_t cgroup_common_file_write(struct cgroup *cont,
+ struct cftype *cft,
+ struct file *file,
+ const char __user *userbuf,
+ size_t nbytes, loff_t *unused_ppos)
+{
+ enum cgroup_filetype type = cft->private;
+ char *buffer;
+ int retval = 0;
+
+ if (nbytes >= PATH_MAX)
+ return -E2BIG;
+
+ /* +1 for nul-terminator */
+ buffer = kmalloc(nbytes + 1, GFP_KERNEL);
+ if (buffer == NULL)
+ return -ENOMEM;
+
+ if (copy_from_user(buffer, userbuf, nbytes)) {
+ retval = -EFAULT;
+ goto out1;
+ }
+ buffer[nbytes] = 0; /* nul-terminate */
+
+ mutex_lock(&cgroup_mutex);
+
+ if (cgroup_is_removed(cont)) {
+ retval = -ENODEV;
+ goto out2;
+ }
+
+ switch (type) {
+ case FILE_TASKLIST:
+ retval = attach_task_by_pid(cont, buffer);
+ break;
+ default:
+ retval = -EINVAL;
+ goto out2;
+ }
+
+ if (retval == 0)
+ retval = nbytes;
+out2:
+ mutex_unlock(&cgroup_mutex);
+out1:
+ kfree(buffer);
+ return retval;
+}
+
static ssize_t cgroup_file_write(struct file *file, const char __user *buf,
size_t nbytes, loff_t *ppos)
{
@@ -924,6 +1094,189 @@ int cgroup_add_files(struct cgroup
return 0;
}
+/* Count the number of tasks in a cgroup. Could be made more
+ * time-efficient but less space-efficient with more linked lists
+ * running through each cgroup and the css_set structures that
+ * referenced it. Must be called with tasklist_lock held for read or
+ * write or in an rcu critical section.
+ */
+int __cgroup_task_count(const struct cgroup *cont)
+{
+ int count = 0;
+ struct task_struct *g, *p;
+ struct cgroup_subsys_state *css;
+ int subsys_id;
+
+ get_first_subsys(cont, &css, &subsys_id);
+ do_each_thread(g, p) {
+ if (task_subsys_state(p, subsys_id) == css)
+ count ++;
+ } while_each_thread(g, p);
+ return count;
+}
+
+/*
+ * Stuff for reading the 'tasks' file.
+ *
+ * Reading this file can return large amounts of data if a cgroup has
+ * *lots* of attached tasks. So it may need several calls to read(),
+ * but we cannot guarantee that the information we produce is correct
+ * unless we produce it entirely atomically.
+ *
+ * Upon tasks file open(), a struct ctr_struct is allocated, that
+ * will have a pointer to an array (also allocated here). The struct
+ * ctr_struct * is stored in file->private_data. Its resources will
+ * be freed by release() when the file is closed. The array is used
+ * to sprintf the PIDs and then used by read().
+ */
+struct ctr_struct {
+ char *buf;
+ int bufsz;
+};
+
+/*
+ * Load into 'pidarray' up to 'npids' of the tasks using cgroup
+ * 'cont'. Return actual number of pids loaded. No need to
+ * task_lock(p) when reading out p->cgroup, since we're in an RCU
+ * read section, so the css_set can't go away, and is
+ * immutable after creation.
+ */
+static int pid_array_load(pid_t *pidarray, int npids, struct cgroup *cont)
+{
+ int n = 0;
+ struct task_struct *g, *p;
+ struct cgroup_subsys_state *css;
+ int subsys_id;
+
+ get_first_subsys(cont, &css, &subsys_id);
+ rcu_read_lock();
+ 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:
+ rcu_read_unlock();
+ return n;
+}
+
+static int cmppid(const void *a, const void *b)
+{
+ return *(pid_t *)a - *(pid_t *)b;
+}
+
+/*
+ * Convert array 'a' of 'npids' pid_t's to a string of newline separated
+ * decimal pids in 'buf'. Don't write more than 'sz' chars, but return
+ * count 'cnt' of how many chars would be written if buf were large enough.
+ */
+static int pid_array_to_buf(char *buf, int sz, pid_t *a, int npids)
+{
+ int cnt = 0;
+ int i;
+
+ for (i = 0; i < npids; i++)
+ cnt += snprintf(buf + cnt, max(sz - cnt, 0), "%d\n", a[i]);
+ return cnt;
+}
+
+/*
+ * Handle an open on 'tasks' file. Prepare a buffer listing the
+ * process id's of tasks currently attached to the cgroup being opened.
+ *
+ * Does not require any specific cgroup mutexes, and does not take any.
+ */
+static int cgroup_tasks_open(struct inode *unused, struct file *file)
+{
+ struct cgroup *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 cgroup 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 cgroup users didn't
+ * show up until sometime later on.
+ */
+ npids = cgroup_task_count(cont);
+ if (npids
...
|
|
|
|
|
|
| [PATCH 04/29] task containersv11 add fork exit hooks [message #20075 is a reply to message #20064] |
Tue, 11 September 2007 19:52 |
Paul Menage
Messages: 642
Registered: September 2006
|
Senior Member |
|
|
From: Paul Menage <menage@google.com>
tAdd adds the necessary hooks to the fork() and exit() paths to ensure that
new children inherit their parent's cgroup assignments, and that exiting
processes release reference counts on their cgroups.
Signed-off-by: Paul Menage <menage@google.com>
Cc: Serge E. Hallyn <serue@us.ibm.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Dave Hansen <haveblue@us.ibm.com>
Cc: Balbir Singh <balbir@in.ibm.com>
Cc: Paul Jackson <pj@sgi.com>
Cc: Kirill Korotaev <dev@openvz.org>
Cc: Herbert Poetzl <herbert@13thfloor.at>
Cc: Srivatsa Vaddagiri <vatsa@in.ibm.com>
Cc: Cedric Le Goater <clg@fr.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---
include/linux/cgroup.h | 6 +
kernel/cgroup.c | 121 ++++++++++++++++++++++++++++++++++++
kernel/exit.c | 2
kernel/fork.c | 14 +++-
4 files changed, 141 insertions(+), 2 deletions(-)
diff -puN include/linux/cgroup.h~task-cgroupsv11-add-fork-exit-hooks include/linux/cgroup.h
--- a/include/linux/cgroup.h~task-cgroupsv11-add-fork-exit-hooks
+++ a/include/linux/cgroup.h
@@ -25,6 +25,9 @@ extern int cgroup_init(void);
extern void cgroup_init_smp(void);
extern void cgroup_lock(void);
extern void cgroup_unlock(void);
+extern void cgroup_fork(struct task_struct *p);
+extern void cgroup_fork_callbacks(struct task_struct *p);
+extern void cgroup_exit(struct task_struct *p, int run_callbacks);
/* Per-subsystem/per-cgroup state maintained by the system. */
struct cgroup_subsys_state {
@@ -215,6 +218,9 @@ int cgroup_path(const struct containe
static inline int cgroup_init_early(void) { return 0; }
static inline int cgroup_init(void) { return 0; }
static inline void cgroup_init_smp(void) {}
+static inline void cgroup_fork(struct task_struct *p) {}
+static inline void cgroup_fork_callbacks(struct task_struct *p) {}
+static inline void cgroup_exit(struct task_struct *p, int callbacks) {}
static inline void cgroup_lock(void) {}
static inline void cgroup_unlock(void) {}
diff -puN kernel/cgroup.c~task-cgroupsv11-add-fork-exit-hooks kernel/cgroup.c
--- a/kernel/cgroup.c~task-cgroupsv11-add-fork-exit-hooks
+++ a/kernel/cgroup.c
@@ -132,6 +132,33 @@ list_for_each_entry(_ss, &_root->subsys_
#define for_each_root(_root) \
list_for_each_entry(_root, &roots, root_list)
+/* Each task_struct has an embedded css_set, so the get/put
+ * operation simply takes a reference count on all the cgroups
+ * referenced by subsystems in this css_set. This can end up
+ * multiple-counting some cgroups, but that's OK - the ref-count is
+ * just a busy/not-busy indicator; ensuring that we only count each
+ * cgroup once would require taking a global lock to ensure that no
+ * subsystems moved between hierarchies while we were doing so.
+ *
+ * Possible TODO: decide at boot time based on the number of
+ * registered subsystems and the number of CPUs or NUMA nodes whether
+ * it's better for performance to ref-count every subsystem, or to
+ * take a global lock and only add one ref count to each hierarchy.
+ */
+static void get_css_set(struct css_set *cg)
+{
+ int i;
+ for (i = 0; i < CGROUP_SUBSYS_COUNT; i++)
+ atomic_inc(&cg->subsys[i]->cgroup->count);
+}
+
+static void put_css_set(struct css_set *cg)
+{
+ int i;
+ for (i = 0; i < CGROUP_SUBSYS_COUNT; i++)
+ atomic_dec(&cg->subsys[i]->cgroup->count);
+}
+
/*
* There is one global cgroup mutex. We also require taking
* task_lock() when dereferencing a task's cgroup subsys pointers.
@@ -1563,3 +1590,97 @@ int __init cgroup_init(void)
out:
return err;
}
+
+/**
+ * cgroup_fork - attach newly forked task to its parents cgroup.
+ * @tsk: pointer to task_struct of forking parent process.
+ *
+ * Description: A task inherits its parent's cgroup at fork().
+ *
+ * A pointer to the shared css_set was automatically copied in
+ * fork.c by dup_task_struct(). However, we ignore that copy, since
+ * it was not made under the protection of RCU or cgroup_mutex, so
+ * might no longer be a valid cgroup pointer. attach_task() might
+ * have already changed current->cgroup, allowing the previously
+ * referenced cgroup to be removed and freed.
+ *
+ * At the point that cgroup_fork() is called, 'current' is the parent
+ * task, and the passed argument 'child' points to the child task.
+ */
+void cgroup_fork(struct task_struct *child)
+{
+ rcu_read_lock();
+ child->cgroups = rcu_dereference(current->cgroups);
+ get_css_set(&child->cgroups);
+ rcu_read_unlock();
+}
+
+/**
+ * cgroup_fork_callbacks - called on a new task very soon before
+ * adding it to the tasklist. No need to take any locks since no-one
+ * can be operating on this task
+ */
+void cgroup_fork_callbacks(struct task_struct *child)
+{
+ if (need_forkexit_callback) {
+ int i;
+ for (i = 0; i < CGROUP_SUBSYS_COUNT; i++) {
+ struct cgroup_subsys *ss = subsys[i];
+ if (ss->fork)
+ ss->fork(ss, child);
+ }
+ }
+}
+
+/**
+ * cgroup_exit - detach cgroup from exiting task
+ * @tsk: pointer to task_struct of exiting process
+ *
+ * Description: Detach cgroup from @tsk and release it.
+ *
+ * Note that cgroups marked notify_on_release force every task in
+ * them to take the global cgroup_mutex mutex when exiting.
+ * This could impact scaling on very large systems. Be reluctant to
+ * use notify_on_release cgroups where very high task exit scaling
+ * is required on large systems.
+ *
+ * the_top_cgroup_hack:
+ *
+ * Set the exiting tasks cgroup to the root cgroup (top_cgroup).
+ *
+ * We call cgroup_exit() while the task is still competent to
+ * handle notify_on_release(), then leave the task attached to the
+ * root cgroup in each hierarchy for the remainder of its exit.
+ *
+ * To do this properly, we would increment the reference count on
+ * top_cgroup, and near the very end of the kernel/exit.c do_exit()
+ * code we would add a second cgroup function call, to drop that
+ * reference. This would just create an unnecessary hot spot on
+ * the top_cgroup reference count, to no avail.
+ *
+ * Normally, holding a reference to a cgroup without bumping its
+ * count is unsafe. The cgroup could go away, or someone could
+ * attach us to a different cgroup, decrementing the count on
+ * the first cgroup that we never incremented. But in this case,
+ * top_cgroup isn't going away, and either task has PF_EXITING set,
+ * which wards off any attach_task() attempts, or task is a failed
+ * fork, never visible to attach_task.
+ *
+ */
+void cgroup_exit(struct task_struct *tsk, int run_callbacks)
+{
+ int i;
+
+ if (run_callbacks && need_forkexit_callback) {
+ for (i = 0; i < CGROUP_SUBSYS_COUNT; i++) {
+ struct cgroup_subsys *ss = subsys[i];
+ if (ss->exit)
+ ss->exit(ss, tsk);
+ }
+ }
+ /* Reassign the task to the init_css_set. */
+ task_lock(tsk);
+ put_css_set(&tsk->cgroups);
+ tsk->cgroups = init_task.cgroups;
+ task_unlock(tsk);
+}
diff -puN kernel/exit.c~task-cgroupsv11-add-fork-exit-hooks kernel/exit.c
--- a/kernel/exit.c~task-cgroupsv11-add-fork-exit-hooks
+++ a/kernel/exit.c
@@ -33,6 +33,7 @@
#include <linux/delayacct.h>
#include <linux/freezer.h>
#include <linux/cpuset.h>
+#include <linux/cgroup.h>
#include <linux/syscalls.h>
#include <linux/signal.h>
#include <linux/posix-timers.h>
@@ -981,6 +982,7 @@ fastcall NORET_TYPE void do_exit(long co
check_stack_usage();
exit_thread();
cpuset_exit(tsk);
+ cgroup_exit(tsk, 1);
exit_keys(tsk);
if (group_dead && tsk->signal->leader)
diff -puN kernel/fork.c~task-cgroupsv11-add-fork-exit-hooks kernel/fork.c
--- a/kernel/fork.c~task-cgroupsv11-add-fork-exit-hooks
+++ a/kernel/fork.c
@@ -30,6 +30,7 @@
#include <linux/capability.h>
#include <linux/cpu.h>
#include <linux/cpuset.h>
+#include <linux/cgroup.h>
#include <linux/security.h>
#include <linux/swap.h>
#include <linux/syscalls.h>
@@ -967,6 +968,7 @@ static struct task_struct *copy_process(
{
int retval;
struct task_struct *p = NULL;
+ int cgroup_callbacks_done = 0;
if ((clone_flags & (CLONE_NEWNS|CLONE_FS)) == (CLONE_NEWNS|CLONE_FS))
return ERR_PTR(-EINVAL);
@@ -1068,12 +1070,13 @@ static struct task_struct *copy_process(
p->io_context = NULL;
p->audit_context = NULL;
cpuset_fork(p);
+ cgroup_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_cgroup;
}
mpol_fix_fork_child_flag(p);
#endif
@@ -1184,6 +1187,12 @@ 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 cgroup callbacks if
+ * necessary. We need to run them before the task is visible
+ * on the tasklist. */
+ cgroup_fork_callbacks(p);
+ cgroup_callbacks_done = 1;
+
/* Need tasklist lock for parent etc handling! */
write_lock_irq(&tasklist_lock);
@@ -1306,9 +1315,10 @@ bad_fork_cleanup_security:
bad_fork_cleanup_policy:
#ifdef CONFIG_NUMA
mpol_free(p->mempolicy);
-bad_fork_cleanup_cpuset:
+bad_fork_cleanup_cgroup:
#endif
cpuset_exit(p);
+ cgroup_exit(p, cgroup_callbacks_done);
delayacct_tsk_free(p);
if (p->binfmt)
module_put(p->binfmt->module);
_
--
_______________________________________________
Containers mailing list
Containers@lists.linux-foundation.org
https://lists.linux-foundation.org/mailman/listinfo/containers
|
|
|
|
| [PATCH 25/29] memory controller add per container lru and reclaim v7 [message #20076 is a reply to message #20064] |
Tue, 11 September 2007 19:53 |
Paul Menage
Messages: 642
Registered: September 2006
|
Senior Member |
|
|
From: Balbir Singh <balbir@linux.vnet.ibm.com>
Add the page_cgroup to the per cgroup LRU. The reclaim algorithm has
been modified to make the isolate_lru_pages() as a pluggable component. The
scan_control data structure now accepts the cgroup on behalf of which
reclaims are carried out. try_to_free_pages() has been extended to become
cgroup aware.
Signed-off-by: Pavel Emelianov <xemul@openvz.org>
Signed-off-by: Balbir Singh <balbir@linux.vnet.ibm.com>
Cc: Paul Menage <menage@google.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Cc: Kirill Korotaev <dev@sw.ru>
Cc: Herbert Poetzl <herbert@13thfloor.at>
Cc: David Rientjes <rientjes@google.com>
Cc: Vaidyanathan Srinivasan <svaidy@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---
include/linux/memcontrol.h | 12 +++
include/linux/res_counter.h | 23 +++++
include/linux/swap.h | 3
mm/memcontrol.c | 135 +++++++++++++++++++++++++++++++++-
mm/swap.c | 2
mm/vmscan.c | 126 +++++++++++++++++++++++++------
6 files changed, 275 insertions(+), 26 deletions(-)
diff -puN include/linux/memcontrol.h~memory-controller-add-per-cgroup-lru-and-reclaim-v7 include/linux/memcontrol.h
--- a/include/linux/memcontrol.h~memory-controller-add-per-cgroup-lru-and-reclaim-v7
+++ a/include/linux/memcontrol.h
@@ -32,6 +32,13 @@ extern void page_assign_page_cgroup(s
extern struct page_cgroup *page_get_page_cgroup(struct page *page);
extern int mem_cgroup_charge(struct page *page, struct mm_struct *mm);
extern void mem_cgroup_uncharge(struct page_cgroup *pc);
+extern void mem_cgroup_move_lists(struct page_cgroup *pc, bool active);
+extern unsigned long mem_cgroup_isolate_pages(unsigned long nr_to_scan,
+ struct list_head *dst,
+ unsigned long *scanned, int order,
+ int mode, struct zone *z,
+ struct mem_cgroup *mem_cont,
+ int active);
static inline void mem_cgroup_uncharge_page(struct page *page)
{
@@ -71,6 +78,11 @@ static inline void mem_cgroup_uncharg
{
}
+static inline void mem_cgroup_move_lists(struct page_cgroup *pc,
+ bool active)
+{
+}
+
#endif /* CONFIG_CGROUP_MEM_CONT */
#endif /* _LINUX_MEMCONTROL_H */
diff -puN include/linux/res_counter.h~memory-controller-add-per-cgroup-lru-and-reclaim-v7 include/linux/res_counter.h
--- a/include/linux/res_counter.h~memory-controller-add-per-cgroup-lru-and-reclaim-v7
+++ a/include/linux/res_counter.h
@@ -99,4 +99,27 @@ int res_counter_charge(struct res_counte
void res_counter_uncharge_locked(struct res_counter *counter, unsigned long val);
void res_counter_uncharge(struct res_counter *counter, unsigned long val);
+static inline bool res_counter_limit_check_locked(struct res_counter *cnt)
+{
+ if (cnt->usage < cnt->limit)
+ return true;
+
+ return false;
+}
+
+/*
+ * Helper function to detect if the cgroup is within it's limit or
+ * not. It's currently called from cgroup_rss_prepare()
+ */
+static inline bool res_counter_check_under_limit(struct res_counter *cnt)
+{
+ bool ret;
+ unsigned long flags;
+
+ spin_lock_irqsave(&cnt->lock, flags);
+ ret = res_counter_limit_check_locked(cnt);
+ spin_unlock_irqrestore(&cnt->lock, flags);
+ return ret;
+}
+
#endif
diff -puN include/linux/swap.h~memory-controller-add-per-cgroup-lru-and-reclaim-v7 include/linux/swap.h
--- a/include/linux/swap.h~memory-controller-add-per-cgroup-lru-and-reclaim-v7
+++ a/include/linux/swap.h
@@ -6,6 +6,7 @@
#include <linux/mmzone.h>
#include <linux/list.h>
#include <linux/sched.h>
+#include <linux/memcontrol.h>
#include <asm/atomic.h>
#include <asm/page.h>
@@ -190,6 +191,8 @@ extern void swap_setup(void);
/* linux/mm/vmscan.c */
extern unsigned long try_to_free_pages(struct zone **zones, int order,
gfp_t gfp_mask);
+extern unsigned long try_to_free_mem_cgroup_pages(struct mem_cgroup *mem);
+extern int __isolate_lru_page(struct page *page, int mode);
extern unsigned long shrink_all_memory(unsigned long nr_pages);
extern int vm_swappiness;
extern int remove_mapping(struct address_space *mapping, struct page *page);
diff -puN mm/memcontrol.c~memory-controller-add-per-cgroup-lru-and-reclaim-v7 mm/memcontrol.c
--- a/mm/memcontrol.c~memory-controller-add-per-cgroup-lru-and-reclaim-v7
+++ a/mm/memcontrol.c
@@ -24,8 +24,12 @@
#include <linux/page-flags.h>
#include <linux/bit_spinlock.h>
#include <linux/rcupdate.h>
+#include <linux/swap.h>
+#include <linux/spinlock.h>
+#include <linux/fs.h>
struct cgroup_subsys mem_cgroup_subsys;
+static const int MEM_CGROUP_RECLAIM_RETRIES = 5;
/*
* The memory controller data structure. The memory controller controls both
@@ -51,6 +55,10 @@ struct mem_cgroup {
*/
struct list_head active_list;
struct list_head inactive_list;
+ /*
+ * spin_lock to protect the per cgroup LRU
+ */
+ spinlock_t lru_lock;
};
/*
@@ -141,6 +149,94 @@ void __always_inline unlock_page_contain
bit_spin_unlock(PAGE_CGROUP_LOCK_BIT, &page->page_cgroup);
}
+void __mem_cgroup_move_lists(struct page_cgroup *pc, bool active)
+{
+ if (active)
+ list_move(&pc->lru, &pc->mem_cgroup->active_list);
+ else
+ list_move(&pc->lru, &pc->mem_cgroup->inactive_list);
+}
+
+/*
+ * This routine assumes that the appropriate zone's lru lock is already held
+ */
+void mem_cgroup_move_lists(struct page_cgroup *pc, bool active)
+{
+ struct mem_cgroup *mem;
+ if (!pc)
+ return;
+
+ mem = pc->mem_cgroup;
+
+ spin_lock(&mem->lru_lock);
+ __mem_cgroup_move_lists(pc, active);
+ spin_unlock(&mem->lru_lock);
+}
+
+unsigned long mem_cgroup_isolate_pages(unsigned long nr_to_scan,
+ struct list_head *dst,
+ unsigned long *scanned, int order,
+ int mode, struct zone *z,
+ struct mem_cgroup *mem_cont,
+ int active)
+{
+ unsigned long nr_taken = 0;
+ struct page *page;
+ unsigned long scan;
+ LIST_HEAD(pc_list);
+ struct list_head *src;
+ struct page_cgroup *pc;
+
+ if (active)
+ src = &mem_cont->active_list;
+ else
+ src = &mem_cont->inactive_list;
+
+ spin_lock(&mem_cont->lru_lock);
+ for (scan = 0; scan < nr_to_scan && !list_empty(src); scan++) {
+ pc = list_entry(src->prev, struct page_cgroup, lru);
+ page = pc->page;
+ VM_BUG_ON(!pc);
+
+ if (PageActive(page) && !active) {
+ __mem_cgroup_move_lists(pc, true);
+ scan--;
+ continue;
+ }
+ if (!PageActive(page) && active) {
+ __mem_cgroup_move_lists(pc, false);
+ scan--;
+ continue;
+ }
+
+ /*
+ * Reclaim, per zone
+ * TODO: make the active/inactive lists per zone
+ */
+ if (page_zone(page) != z)
+ continue;
+
+ /*
+ * Check if the meta page went away from under us
+ */
+ if (!list_empty(&pc->lru))
+ list_move(&pc->lru, &pc_list);
+ else
+ continue;
+
+ if (__isolate_lru_page(page, mode) == 0) {
+ list_move(&page->lru, dst);
+ nr_taken++;
+ }
+ }
+
+ list_splice(&pc_list, src);
+ spin_unlock(&mem_cont->lru_lock);
+
+ *scanned = scan;
+ return nr_taken;
+}
+
/*
* Charge the memory controller for page usage.
* Return
@@ -151,6 +247,8 @@ int mem_cgroup_charge(struct page *pa
{
struct mem_cgroup *mem;
struct page_cgroup *pc, *race_pc;
+ unsigned long flags;
+ unsigned long nr_retries = MEM_CGROUP_RECLAIM_RETRIES;
/*
* Should page_cgroup's go to their own slab?
@@ -197,7 +295,32 @@ int mem_cgroup_charge(struct page *pa
* If we created the page_cgroup, we should free it on exceeding
* the cgroup limit.
*/
- if (res_counter_charge(&mem->res, 1)) {
+ while (res_counter_charge(&mem->res, 1)) {
+ if (try_to_free_mem_cgroup_pages(mem))
+ continue;
+
+ /*
+ * try_to_free_mem_cgroup_pages() might not give us a full
+ * picture of reclaim. Some pages are reclaimed and might be
+ * moved to swap cache or just unmapped from the cgroup.
+ * Check the limit again to see if the reclaim reduced the
+ * current usage of the cgroup before giving up
+ */
+ if (res_counter_check_under_limit(&mem->res))
+ continue;
+ /*
+ * Since we control both RSS and cache, we end up with a
+ * very interesting scenario where we end up reclaiming
+ * memory (essentially RSS), since the memory is pushed
+ * to swap cache, we eventually end up adding those
+ * pages back to our list. Hence we give ourselves a
+ * few chances before we fail
+ */
+ else if (nr_retries--) {
+ congestion_wait(WRITE, HZ/10);
+ continue;
+ }
+
css_put(&mem->css);
goto free_pc;
}
@@ -221,6 +344,10 @@ int mem_cgroup_charge(struct page *pa
pc->page = page;
page_assign_page_cgroup(page, pc);
+ spin_lock_irqsave(&mem->lru_lock, flags);
+ list_add(&pc->lru, &mem->active_list);
+ spin_unlock_irqrestore(&mem->lru_lock, flags);
+
done:
unlock_page_cgroup(page);
return 0;
@@ -240,6 +367,7 @@ void mem_cgroup_uncharge(struct page_
{
struct mem_cgroup *mem;
struct page *page;
+ unsigned long flags;
if (!pc)
return;
@@ -252,6 +380,10 @@ void mem_cgroup_uncharge(struct page_
page_assign_page_cgroup(page, NULL);
unlock_page_cgroup(page);
res_counter_uncharge(&mem->res, 1);
+
+ spin_lock_irqsave(&mem->lru_lock, flags);
+ list_del_init(&pc->lru);
+ spin_unlock_irqrestore(&mem->lru_lock, flags);
kfree(pc);
}
}
@@ -310,6 +442,7 @@ mem_cgroup_create(struct cgroup_su
res_counter_init(&mem->res);
INIT_LIST_HEAD(&mem->active_list);
INIT_LIST_HEAD(&mem->inactive_list);
+ spin_lock_init(&mem->lru_lock);
return &mem->css;
}
diff -puN mm/swap.c~memory-controller-add-per-cgroup-lru-and-reclaim-v7 mm/swap.c
--- a/mm/swap.
...
|
|
|
|
|
|
| [PATCH 23/29] memory controller memory accounting v7 [message #20078 is a reply to message #20064] |
Tue, 11 September 2007 19:53 |
Paul Menage
Messages: 642
Registered: September 2006
|
Senior Member |
|
|
From: Balbir Singh <balbir@linux.vnet.ibm.com>
Add the accounting hooks. The accounting is carried out for RSS and Page
Cache (unmapped) pages. There is now a common limit and accounting for both.
The RSS accounting is accounted at page_add_*_rmap() and page_remove_rmap()
time. Page cache is accounted at add_to_page_cache(),
__delete_from_page_cache(). Swap cache is also accounted for.
Each page's page_cgroup is protected with the last bit of the
page_cgroup pointer, this makes handling of race conditions involving
simultaneous mappings of a page easier. A reference count is kept in the
page_cgroup to deal with cases where a page might be unmapped from the RSS
of all tasks, but still lives in the page cache.
Credits go to Vaidyanathan Srinivasan for helping with reference counting work
of the page cgroup. Almost all of the page cache accounting code has help
from Vaidyanathan Srinivasan.
Signed-off-by: Vaidyanathan Srinivasan <svaidy@linux.vnet.ibm.com>
Signed-off-by: Balbir Singh <balbir@linux.vnet.ibm.com>
Cc: Pavel Emelianov <xemul@openvz.org>
Cc: Paul Menage <menage@google.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Cc: Kirill Korotaev <dev@sw.ru>
Cc: Herbert Poetzl <herbert@13thfloor.at>
Cc: David Rientjes <rientjes@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---
include/linux/memcontrol.h | 20 ++++
mm/filemap.c | 12 ++
mm/memcontrol.c | 166 ++++++++++++++++++++++++++++++++++-
mm/memory.c | 43 ++++++++-
mm/migrate.c | 6 +
mm/page_alloc.c | 3
mm/rmap.c | 17 +++
mm/swap_state.c | 12 ++
mm/swapfile.c | 41 +++++---
9 files changed, 293 insertions(+), 27 deletions(-)
diff -puN include/linux/memcontrol.h~memory-controller-memory-accounting-v7 include/linux/memcontrol.h
--- a/include/linux/memcontrol.h~memory-controller-memory-accounting-v7
+++ a/include/linux/memcontrol.h
@@ -30,6 +30,13 @@ extern void mm_free_cgroup(struct mm_
extern void page_assign_page_cgroup(struct page *page,
struct page_cgroup *pc);
extern struct page_cgroup *page_get_page_cgroup(struct page *page);
+extern int mem_cgroup_charge(struct page *page, struct mm_struct *mm);
+extern void mem_cgroup_uncharge(struct page_cgroup *pc);
+
+static inline void mem_cgroup_uncharge_page(struct page *page)
+{
+ mem_cgroup_uncharge(page_get_page_cgroup(page));
+}
#else /* CONFIG_CGROUP_MEM_CONT */
static inline void mm_init_cgroup(struct mm_struct *mm,
@@ -51,6 +58,19 @@ static inline struct page_cgroup *pag
return NULL;
}
+static inline int mem_cgroup_charge(struct page *page, struct mm_struct *mm)
+{
+ return 0;
+}
+
+static inline void mem_cgroup_uncharge(struct page_cgroup *pc)
+{
+}
+
+static inline void mem_cgroup_uncharge_page(struct page *page)
+{
+}
+
#endif /* CONFIG_CGROUP_MEM_CONT */
#endif /* _LINUX_MEMCONTROL_H */
diff -puN mm/filemap.c~memory-controller-memory-accounting-v7 mm/filemap.c
--- a/mm/filemap.c~memory-controller-memory-accounting-v7
+++ a/mm/filemap.c
@@ -31,6 +31,7 @@
#include <linux/syscalls.h>
#include <linux/cpuset.h>
#include <linux/hardirq.h> /* for BUG_ON(!in_atomic()) only */
+#include <linux/memcontrol.h>
#include "internal.h"
/*
@@ -116,6 +117,7 @@ void __remove_from_page_cache(struct pag
{
struct address_space *mapping = page->mapping;
+ mem_cgroup_uncharge_page(page);
radix_tree_delete(&mapping->page_tree, page->index);
page->mapping = NULL;
mapping->nrpages--;
@@ -441,6 +443,11 @@ int add_to_page_cache(struct page *page,
int error = radix_tree_preload(gfp_mask & ~__GFP_HIGHMEM);
if (error == 0) {
+
+ error = mem_cgroup_charge(page, current->mm);
+ if (error)
+ goto out;
+
write_lock_irq(&mapping->tree_lock);
error = radix_tree_insert(&mapping->page_tree, offset, page);
if (!error) {
@@ -450,10 +457,13 @@ int add_to_page_cache(struct page *page,
page->index = offset;
mapping->nrpages++;
__inc_zone_page_state(page, NR_FILE_PAGES);
- }
+ } else
+ mem_cgroup_uncharge_page(page);
+
write_unlock_irq(&mapping->tree_lock);
radix_tree_preload_end();
}
+out:
return error;
}
EXPORT_SYMBOL(add_to_page_cache);
diff -puN mm/memcontrol.c~memory-controller-memory-accounting-v7 mm/memcontrol.c
--- a/mm/memcontrol.c~memory-controller-memory-accounting-v7
+++ a/mm/memcontrol.c
@@ -21,6 +21,9 @@
#include <linux/memcontrol.h>
#include <linux/cgroup.h>
#include <linux/mm.h>
+#include <linux/page-flags.h>
+#include <linux/bit_spinlock.h>
+#include <linux/rcupdate.h>
struct cgroup_subsys mem_cgroup_subsys;
@@ -31,7 +34,9 @@ struct cgroup_subsys mem_cgroup_su
* to help the administrator determine what knobs to tune.
*
* TODO: Add a water mark for the memory controller. Reclaim will begin when
- * we hit the water mark.
+ * we hit the water mark. May be even add a low water mark, such that
+ * no reclaim occurs from a cgroup at it's low water mark, this is
+ * a feature that will be implemented much later in the future.
*/
struct mem_cgroup {
struct cgroup_subsys_state css;
@@ -49,6 +54,14 @@ struct mem_cgroup {
};
/*
+ * We use the lower bit of the page->page_cgroup pointer as a bit spin
+ * lock. We need to ensure that page->page_cgroup is atleast two
+ * byte aligned (based on comments from Nick Piggin)
+ */
+#define PAGE_CGROUP_LOCK_BIT 0x0
+#define PAGE_CGROUP_LOCK (1 << PAGE_CGROUP_LOCK_BIT)
+
+/*
* A page_cgroup page is associated with every page descriptor. The
* page_cgroup helps us identify information about the cgroup
*/
@@ -56,6 +69,8 @@ struct page_cgroup {
struct list_head lru; /* per cgroup LRU list */
struct page *page;
struct mem_cgroup *mem_cgroup;
+ atomic_t ref_cnt; /* Helpful when pages move b/w */
+ /* mapped and cached states */
};
@@ -88,14 +103,157 @@ void mm_free_cgroup(struct mm_struct
css_put(&mm->mem_cgroup->css);
}
+static inline int page_cgroup_locked(struct page *page)
+{
+ return bit_spin_is_locked(PAGE_CGROUP_LOCK_BIT,
+ &page->page_cgroup);
+}
+
void page_assign_page_cgroup(struct page *page, struct page_cgroup *pc)
{
- page->page_cgroup = (unsigned long)pc;
+ int locked;
+
+ /*
+ * While resetting the page_cgroup we might not hold the
+ * page_cgroup lock. free_hot_cold_page() is an example
+ * of such a scenario
+ */
+ if (pc)
+ VM_BUG_ON(!page_cgroup_locked(page));
+ locked = (page->page_cgroup & PAGE_CGROUP_LOCK);
+ page->page_cgroup = ((unsigned long)pc | locked);
}
struct page_cgroup *page_get_page_cgroup(struct page *page)
{
- return page->page_cgroup;
+ return (struct page_cgroup *)
+ (page->page_cgroup & ~PAGE_CGROUP_LOCK);
+}
+
+void __always_inline lock_page_cgroup(struct page *page)
+{
+ bit_spin_lock(PAGE_CGROUP_LOCK_BIT, &page->page_cgroup);
+ VM_BUG_ON(!page_cgroup_locked(page));
+}
+
+void __always_inline unlock_page_cgroup(struct page *page)
+{
+ bit_spin_unlock(PAGE_CGROUP_LOCK_BIT, &page->page_cgroup);
+}
+
+/*
+ * Charge the memory controller for page usage.
+ * Return
+ * 0 if the charge was successful
+ * < 0 if the cgroup is over its limit
+ */
+int mem_cgroup_charge(struct page *page, struct mm_struct *mm)
+{
+ struct mem_cgroup *mem;
+ struct page_cgroup *pc, *race_pc;
+
+ /*
+ * Should page_cgroup's go to their own slab?
+ * One could optimize the performance of the charging routine
+ * by saving a bit in the page_flags and using it as a lock
+ * to see if the cgroup page already has a page_cgroup associated
+ * with it
+ */
+ lock_page_cgroup(page);
+ pc = page_get_page_cgroup(page);
+ /*
+ * The page_cgroup exists and the page has already been accounted
+ */
+ if (pc) {
+ atomic_inc(&pc->ref_cnt);
+ goto done;
+ }
+
+ unlock_page_cgroup(page);
+
+ pc = kzalloc(sizeof(struct page_cgroup), GFP_KERNEL);
+ if (pc == NULL)
+ goto err;
+
+ rcu_read_lock();
+ /*
+ * We always charge the cgroup the mm_struct belongs to
+ * the mm_struct's mem_cgroup changes on task migration if the
+ * thread group leader migrates. It's possible that mm is not
+ * set, if so charge the init_mm (happens for pagecache usage).
+ */
+ if (!mm)
+ mm = &init_mm;
+
+ mem = rcu_dereference(mm->mem_cgroup);
+ /*
+ * For every charge from the cgroup, increment reference
+ * count
+ */
+ css_get(&mem->css);
+ rcu_read_unlock();
+
+ /*
+ * If we created the page_cgroup, we should free it on exceeding
+ * the cgroup limit.
+ */
+ if (res_counter_charge(&mem->res, 1)) {
+ css_put(&mem->css);
+ goto free_pc;
+ }
+
+ lock_page_cgroup(page);
+ /*
+ * Check if somebody else beat us to allocating the page_cgroup
+ */
+ race_pc = page_get_page_cgroup(page);
+ if (race_pc) {
+ kfree(pc);
+ pc = race_pc;
+ atomic_inc(&pc->ref_cnt);
+ res_counter_uncharge(&mem->res, 1);
+ css_put(&mem->css);
+ goto done;
+ }
+
+ atomic_set(&pc->ref_cnt, 1);
+ pc->mem_cgroup = mem;
+ pc->page = page;
+ page_assign_page_cgroup(page, pc);
+
+done:
+ unlock_page_cgroup(page);
+ return 0;
+free_pc:
+ kfree(pc);
+ return -ENOMEM;
+err:
+ unlock_page_cgroup(page);
+ return -ENOMEM;
+}
+
+/*
+ * Uncharging is always a welcome operation, we never complain, simply
+ * uncharge.
+ */
+void mem_cgroup_uncharge(struct page_cgroup *pc)
+{
+ struct mem_cgroup *mem;
+ struct page *page;
+
+ if (!pc)
+ return;
+
+ if (atomic_dec_and_test(&pc->ref_cnt)) {
+ page = pc->page;
+ lock_page_cgroup(page);
+ mem = pc->mem_cgroup;
+ css_put(&mem->css);
+ page_assign_page_cgroup(page, NULL);
+ unlock_page_cgroup(page);
+ res_counter_uncharge(&m
...
|
|
|
|
| [PATCH 05/29] task containersv11 add container_clone interface [message #20081 is a reply to message #20064] |
Tue, 11 September 2007 19:52 |
Paul Menage
Messages: 642
Registered: September 2006
|
Senior Member |
|
|
From: Paul Menage <menage@google.com>
Add support for cgroup_clone(), a way to create new cgroups intended to
be used for systems such as namespace unsharing. A new subsystem callback,
post_clone(), is added to allow subsystems to automatically configure cloned
cgroups.
Signed-off-by: Paul Menage <menage@google.com>
Cc: Serge E. Hallyn <serue@us.ibm.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Dave Hansen <haveblue@us.ibm.com>
Cc: Balbir Singh <balbir@in.ibm.com>
Cc: Paul Jackson <pj@sgi.com>
Cc: Kirill Korotaev <dev@openvz.org>
Cc: Herbert Poetzl <herbert@13thfloor.at>
Cc: Srivatsa Vaddagiri <vatsa@in.ibm.com>
Cc: Cedric Le Goater <clg@fr.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---
Documentation/cgroups.txt | 7 +
include/linux/cgroup.h | 3
kernel/cgroup.c | 135 +++++++++++++++++++++++++++++++++
3 files changed, 145 insertions(+)
diff -puN Documentation/cgroups.txt~task-cgroupsv11-add-cgroup_clone-interface Documentation/cgroups.txt
--- a/Documentation/cgroups.txt~task-cgroupsv11-add-cgroup_clone-interface
+++ a/Documentation/cgroups.txt
@@ -504,6 +504,13 @@ include/linux/cgroup.h for details).
method can return an error code, the error code is currently not
always handled well.
+void post_clone(struct cgroup_subsys *ss, struct cgroup *cont)
+
+Called at the end of cgroup_clone() to do any paramater
+initialization which might be required before a task could attach. For
+example in cpusets, no task may attach before 'cpus' and 'mems' are set
+up.
+
void bind(struct cgroup_subsys *ss, struct cgroup *root)
LL=callback_mutex
diff -puN include/linux/cgroup.h~task-cgroupsv11-add-cgroup_clone-interface include/linux/cgroup.h
--- a/include/linux/cgroup.h~task-cgroupsv11-add-cgroup_clone-interface
+++ a/include/linux/cgroup.h
@@ -174,6 +174,7 @@ struct cgroup_subsys {
void (*exit)(struct cgroup_subsys *ss, struct task_struct *task);
int (*populate)(struct cgroup_subsys *ss,
struct cgroup *cont);
+ void (*post_clone)(struct cgroup_subsys *ss, struct cgroup *cont);
void (*bind)(struct cgroup_subsys *ss, struct cgroup *root);
int subsys_id;
int active;
@@ -213,6 +214,8 @@ static inline struct cgroup* task_con
int cgroup_path(const struct cgroup *cont, char *buf, int buflen);
+int cgroup_clone(struct task_struct *tsk, struct cgroup_subsys *ss);
+
#else /* !CONFIG_CGROUPS */
static inline int cgroup_init_early(void) { return 0; }
diff -puN kernel/cgroup.c~task-cgroupsv11-add-cgroup_clone-interface kernel/cgroup.c
--- a/kernel/cgroup.c~task-cgroupsv11-add-cgroup_clone-interface
+++ a/kernel/cgroup.c
@@ -1684,3 +1684,138 @@ void cgroup_exit(struct task_struct *
tsk->cgroups = init_task.cgroups;
task_unlock(tsk);
}
+
+/**
+ * cgroup_clone - duplicate the current cgroup in the hierarchy
+ * that the given subsystem is attached to, and move this task into
+ * the new child
+ */
+int cgroup_clone(struct task_struct *tsk, struct cgroup_subsys *subsys)
+{
+ struct dentry *dentry;
+ int ret = 0;
+ char nodename[MAX_CGROUP_TYPE_NAMELEN];
+ struct cgroup *parent, *child;
+ struct inode *inode;
+ struct css_set *cg;
+ struct cgroupfs_root *root;
+ struct cgroup_subsys *ss;
+
+ /* We shouldn't be called by an unregistered subsystem */
+ BUG_ON(!subsys->active);
+
+ /* First figure out what hierarchy and cgroup we're dealing
+ * with, and pin them so we can drop cgroup_mutex */
+ mutex_lock(&cgroup_mutex);
+ again:
+ root = subsys->root;
+ if (root == &rootnode) {
+ printk(KERN_INFO
+ "Not cloning cgroup for unused subsystem %s\n",
+ subsys->name);
+ mutex_unlock(&cgroup_mutex);
+ return 0;
+ }
+ cg = &tsk->cgroups;
+ parent = task_cgroup(tsk, subsys->subsys_id);
+
+ snprintf(nodename, MAX_CGROUP_TYPE_NAMELEN, "node_%d", tsk->pid);
+
+ /* Pin the hierarchy */
+ atomic_inc(&parent->root->sb->s_active);
+
+ mutex_unlock(&cgroup_mutex);
+
+ /* Now do the VFS work to create a cgroup */
+ inode = parent->dentry->d_inode;
+
+ /* Hold the parent directory mutex across this operation to
+ * stop anyone else deleting the new cgroup */
+ mutex_lock(&inode->i_mutex);
+ dentry = cgroup_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 cgroup directory, which also creates the cgroup */
+ ret = vfs_mkdir(inode, dentry, S_IFDIR | 0755);
+ child = __d_cont(dentry);
+ dput(dentry);
+ if (ret) {
+ printk(KERN_INFO
+ "Failed to create cgroup %s: %d\n", nodename,
+ ret);
+ goto out_release;
+ }
+
+ if (!child) {
+ printk(KERN_INFO
+ "Couldn't find new cgroup %s\n", nodename);
+ ret = -ENOMEM;
+ goto out_release;
+ }
+
+ /* The cgroup now exists. Retake cgroup_mutex and check
+ * that we're still in the same state that we thought we
+ * were. */
+ mutex_lock(&cgroup_mutex);
+ if ((root != subsys->root) ||
+ (parent != task_cgroup(tsk, subsys->subsys_id))) {
+ /* Aargh, we raced ... */
+ mutex_unlock(&inode->i_mutex);
+
+ deactivate_super(parent->root->sb);
+ /* The cgroup is still accessible in the VFS, but
+ * we're not going to try to rmdir() it at this
+ * point. */
+ printk(KERN_INFO
+ "Race in cgroup_clone() - leaking cgroup %s\n",
+ nodename);
+ goto again;
+ }
+
+ /* do any required auto-setup */
+ for_each_subsys(root, ss) {
+ if (ss->post_clone)
+ ss->post_clone(ss, child);
+ }
+
+ /* All seems fine. Finish by moving the task into the new cgroup */
+ ret = attach_task(child, tsk);
+ mutex_unlock(&cgroup_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 cgroup in
+ * the appropriate hierarchy
+ *
+ * If we are sending in dummytop, then presumably we are creating
+ * the top cgroup in the subsystem.
+ *
+ * Called only by the ns (nsproxy) cgroup.
+ */
+int cgroup_is_descendant(const struct cgroup *cont)
+{
+ int ret;
+ struct cgroup *target;
+ int subsys_id;
+
+ if (cont == dummytop)
+ return 1;
+
+ get_first_subsys(cont, NULL, &subsys_id);
+ target = task_cgroup(current, subsys_id);
+ while (cont != target && cont!= cont->top_cgroup)
+ cont = cont->parent;
+ ret = (cont == target);
+ return ret;
+}
_
--
_______________________________________________
Containers mailing list
Containers@lists.linux-foundation.org
https://lists.linux-foundation.org/mailman/listinfo/containers
|
|
|
|
| [PATCH 29/29] memory controller make page_referenced container aware v7 [message #20082 is a reply to message #20064] |
Tue, 11 September 2007 19:53 |
Paul Menage
Messages: 642
Registered: September 2006
|
Senior Member |
|
|
From: Balbir Singh <balbir@linux.vnet.ibm.com>
Make page_referenced() cgroup aware. Without this patch, page_referenced()
can cause a page to be skipped while reclaiming pages. This patch ensures
that other cgroups do not hold pages in a particular cgroup hostage. It
is required to ensure that shared pages are freed from a cgroup when they
are not actively referenced from the cgroup that brought them in
Signed-off-by: Balbir Singh <balbir@linux.vnet.ibm.com>
Cc: Pavel Emelianov <xemul@openvz.org>
Cc: Paul Menage <menage@google.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Cc: Kirill Korotaev <dev@sw.ru>
Cc: Herbert Poetzl <herbert@13thfloor.at>
Cc: David Rientjes <rientjes@google.com>
Cc: Vaidyanathan Srinivasan <svaidy@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---
include/linux/memcontrol.h | 6 ++++++
include/linux/rmap.h | 5 +++--
mm/memcontrol.c | 5 +++++
mm/rmap.c | 30 ++++++++++++++++++++++++------
mm/vmscan.c | 4 ++--
5 files changed, 40 insertions(+), 10 deletions(-)
diff -puN include/linux/memcontrol.h~memory-controller-make-page_referenced-cgroup-aware-v7 include/linux/memcontrol.h
--- a/include/linux/memcontrol.h~memory-controller-make-page_referenced-cgroup-aware-v7
+++ a/include/linux/memcontrol.h
@@ -43,6 +43,7 @@ extern unsigned long mem_cgroup_isola
int active);
extern void mem_cgroup_out_of_memory(struct mem_cgroup *mem);
extern int mem_cgroup_cache_charge(struct page *page, struct mm_struct *mm);
+extern struct mem_cgroup *mm_cgroup(struct mm_struct *mm);
static inline void mem_cgroup_uncharge_page(struct page *page)
{
@@ -93,6 +94,11 @@ static inline int mem_cgroup_cache_ch
return 0;
}
+static inline struct mem_cgroup *mm_cgroup(struct mm_struct *mm)
+{
+ return NULL;
+}
+
#endif /* CONFIG_CGROUP_MEM_CONT */
#endif /* _LINUX_MEMCONTROL_H */
diff -puN include/linux/rmap.h~memory-controller-make-page_referenced-cgroup-aware-v7 include/linux/rmap.h
--- a/include/linux/rmap.h~memory-controller-make-page_referenced-cgroup-aware-v7
+++ a/include/linux/rmap.h
@@ -8,6 +8,7 @@
#include <linux/slab.h>
#include <linux/mm.h>
#include <linux/spinlock.h>
+#include <linux/memcontrol.h>
/*
* The anon_vma heads a list of private "related" vmas, to scan if
@@ -86,7 +87,7 @@ static inline void page_dup_rmap(struct
/*
* Called from mm/vmscan.c to handle paging out
*/
-int page_referenced(struct page *, int is_locked);
+int page_referenced(struct page *, int is_locked, struct mem_cgroup *cnt);
int try_to_unmap(struct page *, int ignore_refs);
/*
@@ -114,7 +115,7 @@ int page_mkclean(struct page *);
#define anon_vma_prepare(vma) (0)
#define anon_vma_link(vma) do {} while (0)
-#define page_referenced(page,l) TestClearPageReferenced(page)
+#define page_referenced(page,l,cnt) TestClearPageReferenced(page)
#define try_to_unmap(page, refs) SWAP_FAIL
static inline int page_mkclean(struct page *page)
diff -puN mm/memcontrol.c~memory-controller-make-page_referenced-cgroup-aware-v7 mm/memcontrol.c
--- a/mm/memcontrol.c~memory-controller-make-page_referenced-cgroup-aware-v7
+++ a/mm/memcontrol.c
@@ -109,6 +109,11 @@ struct mem_cgroup *mem_cgroup_from
struct mem_cgroup, css);
}
+inline struct mem_cgroup *mm_cgroup(struct mm_struct *mm)
+{
+ return rcu_dereference(mm->mem_cgroup);
+}
+
void mm_init_cgroup(struct mm_struct *mm, struct task_struct *p)
{
struct mem_cgroup *mem;
diff -puN mm/rmap.c~memory-controller-make-page_referenced-cgroup-aware-v7 mm/rmap.c
--- a/mm/rmap.c~memory-controller-make-page_referenced-cgroup-aware-v7
+++ a/mm/rmap.c
@@ -299,7 +299,8 @@ out:
return referenced;
}
-static int page_referenced_anon(struct page *page)
+static int page_referenced_anon(struct page *page,
+ struct mem_cgroup *mem_cont)
{
unsigned int mapcount;
struct anon_vma *anon_vma;
@@ -312,6 +313,13 @@ static int page_referenced_anon(struct p
mapcount = page_mapcount(page);
list_for_each_entry(vma, &anon_vma->head, anon_vma_node) {
+ /*
+ * If we are reclaiming on behalf of a cgroup, skip
+ * counting on behalf of references from different
+ * cgroups
+ */
+ if (mem_cont && (mm_cgroup(vma->vm_mm) != mem_cont))
+ continue;
referenced += page_referenced_one(page, vma, &mapcount);
if (!mapcount)
break;
@@ -332,7 +340,8 @@ static int page_referenced_anon(struct p
*
* This function is only called from page_referenced for object-based pages.
*/
-static int page_referenced_file(struct page *page)
+static int page_referenced_file(struct page *page,
+ struct mem_cgroup *mem_cont)
{
unsigned int mapcount;
struct address_space *mapping = page->mapping;
@@ -365,6 +374,13 @@ static int page_referenced_file(struct p
mapcount = page_mapcount(page);
vma_prio_tree_foreach(vma, &iter, &mapping->i_mmap, pgoff, pgoff) {
+ /*
+ * If we are reclaiming on behalf of a cgroup, skip
+ * counting on behalf of references from different
+ * cgroups
+ */
+ if (mem_cont && (mm_cgroup(vma->vm_mm) != mem_cont))
+ continue;
if ((vma->vm_flags & (VM_LOCKED|VM_MAYSHARE))
== (VM_LOCKED|VM_MAYSHARE)) {
referenced++;
@@ -387,7 +403,8 @@ static int page_referenced_file(struct p
* Quick test_and_clear_referenced for all mappings to a page,
* returns the number of ptes which referenced the page.
*/
-int page_referenced(struct page *page, int is_locked)
+int page_referenced(struct page *page, int is_locked,
+ struct mem_cgroup *mem_cont)
{
int referenced = 0;
@@ -399,14 +416,15 @@ int page_referenced(struct page *page, i
if (page_mapped(page) && page->mapping) {
if (PageAnon(page))
- referenced += page_referenced_anon(page);
+ referenced += page_referenced_anon(page, mem_cont);
else if (is_locked)
- referenced += page_referenced_file(page);
+ referenced += page_referenced_file(page, mem_cont);
else if (TestSetPageLocked(page))
referenced++;
else {
if (page->mapping)
- referenced += page_referenced_file(page);
+ referenced +=
+ page_referenced_file(page, mem_cont);
unlock_page(page);
}
}
diff -puN mm/vmscan.c~memory-controller-make-page_referenced-cgroup-aware-v7 mm/vmscan.c
--- a/mm/vmscan.c~memory-controller-make-page_referenced-cgroup-aware-v7
+++ a/mm/vmscan.c
@@ -503,7 +503,7 @@ static unsigned long shrink_page_list(st
goto keep_locked;
}
- referenced = page_referenced(page, 1);
+ referenced = page_referenced(page, 1, sc->mem_cgroup);
/* In active use or really unfreeable? Activate it. */
if (sc->order <= PAGE_ALLOC_COSTLY_ORDER &&
referenced && page_mapping_inuse(page))
@@ -1057,7 +1057,7 @@ force_reclaim_mapped:
if (page_mapped(page)) {
if (!reclaim_mapped ||
(total_swap_pages == 0 && PageAnon(page)) ||
- page_referenced(page, 0)) {
+ page_referenced(page, 0, sc->mem_cgroup)) {
list_add(&page->lru, &l_active);
continue;
}
_
--
_______________________________________________
Containers mailing list
Containers@lists.linux-foundation.org
https://lists.linux-foundation.org/mailman/listinfo/containers
|
|
|
|
| [PATCH 10/29] task containersv11 automatic userspace notification of idle containers [message #20083 is a reply to message #20064] |
Tue, 11 September 2007 19:52 |
Paul Menage
Messages: 642
Registered: September 2006
|
Senior Member |
|
|
From: Paul Menage <menage@google.com>
Add the following files to the cgroup filesystem:
notify_on_release - configures/reports whether the cgroup subsystem should
attempt to run a release script when this cgroup becomes unused
release_agent - configures/reports the release agent to be used for this
hierarchy (top level in each hierarchy only)
releasable - reports whether this cgroup would have been auto-released if
notify_on_release was true and a release agent was configured (mainly useful
for debugging)
To avoid locking issues, invoking the userspace release agent is done via a
workqueue task; cgroups that need to have their release agents invoked by
the workqueue task are linked on to a list.
Signed-off-by: Paul Menage <menage@google.com>
Cc: Serge E. Hallyn <serue@us.ibm.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Dave Hansen <haveblue@us.ibm.com>
Cc: Balbir Singh <balbir@in.ibm.com>
Cc: Paul Jackson <pj@sgi.com>
Cc: Kirill Korotaev <dev@openvz.org>
Cc: Herbert Poetzl <herbert@13thfloor.at>
Cc: Srivatsa Vaddagiri <vatsa@in.ibm.com>
Cc: Cedric Le Goater <clg@fr.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---
include/linux/cgroup.h | 11
kernel/cgroup.c | 425 ++++++++++++++++++++++++++++++++----
2 files changed, 393 insertions(+), 43 deletions(-)
diff -puN include/linux/cgroup.h~task-cgroupsv11-automatic-userspace-notification-of-idle-cgroups include/linux/cgroup.h
--- a/include/linux/cgroup.h~task-cgroupsv11-automatic-userspace-notification-of-idle-cgroups
+++ a/include/linux/cgroup.h
@@ -77,10 +77,11 @@ static inline void css_get(struct contai
* css_get()
*/
+extern void __css_put(struct cgroup_subsys_state *css);
static inline void css_put(struct cgroup_subsys_state *css)
{
if (!test_bit(CSS_ROOT, &css->flags))
- atomic_dec(&css->refcnt);
+ __css_put(css);
}
struct cgroup {
@@ -112,6 +113,13 @@ struct cgroup {
* tasks in this cgroup. Protected by css_set_lock
*/
struct list_head css_sets;
+
+ /*
+ * Linked list running through all cgroups that can
+ * potentially be reaped by the release agent. Protected by
+ * release_list_lock
+ */
+ struct list_head release_list;
};
/* A css_set is a structure holding pointers to a set of
@@ -285,7 +293,6 @@ struct task_struct *cgroup_iter_next(
struct cgroup_iter *it);
void cgroup_iter_end(struct cgroup *cont, struct cgroup_iter *it);
-
#else /* !CONFIG_CGROUPS */
static inline int cgroup_init_early(void) { return 0; }
diff -puN kernel/cgroup.c~task-cgroupsv11-automatic-userspace-notification-of-idle-cgroups kernel/cgroup.c
--- a/kernel/cgroup.c~task-cgroupsv11-automatic-userspace-notification-of-idle-cgroups
+++ a/kernel/cgroup.c
@@ -44,6 +44,8 @@
#include <linux/sort.h>
#include <asm/atomic.h>
+static DEFINE_MUTEX(cgroup_mutex);
+
/* Generate an array of cgroup subsystem pointers */
#define SUBSYS(_x) &_x ## _subsys,
@@ -82,6 +84,13 @@ struct cgroupfs_root {
/* Hierarchy-specific flags */
unsigned long flags;
+
+ /* The path to use for release notifications. No locking
+ * between setting and use - so if userspace updates this
+ * while child cgroups exist, you could miss a
+ * notification. We ensure that it's always a valid
+ * NUL-terminated string */
+ char release_agent_path[PATH_MAX];
};
@@ -109,7 +118,13 @@ static int need_forkexit_callback;
/* bits in struct cgroup flags field */
enum {
+ /* Control Group is dead */
CONT_REMOVED,
+ /* Control Group has previously had a child cgroup or a task,
+ * but no longer (only if CONT_NOTIFY_ON_RELEASE is set) */
+ CONT_RELEASABLE,
+ /* Control Group requires release notifications to userspace */
+ CONT_NOTIFY_ON_RELEASE,
};
/* convenient tests for these bits */
@@ -123,6 +138,19 @@ enum {
ROOT_NOPREFIX, /* mounted subsystems have no named prefix */
};
+inline int cgroup_is_releasable(const struct cgroup *cont)
+{
+ const int bits =
+ (1 << CONT_RELEASABLE) |
+ (1 << CONT_NOTIFY_ON_RELEASE);
+ return (cont->flags & bits) == bits;
+}
+
+inline int notify_on_release(const struct cgroup *cont)
+{
+ return test_bit(CONT_NOTIFY_ON_RELEASE, &cont->flags);
+}
+
/*
* for_each_subsys() allows you to iterate on each subsystem attached to
* an active hierarchy
@@ -134,6 +162,14 @@ list_for_each_entry(_ss, &_root->subsys_
#define for_each_root(_root) \
list_for_each_entry(_root, &roots, root_list)
+/* the list of cgroups eligible for automatic release. Protected by
+ * release_list_lock */
+static LIST_HEAD(release_list);
+static DEFINE_SPINLOCK(release_list_lock);
+static void cgroup_release_agent(struct work_struct *work);
+static DECLARE_WORK(release_agent_work, cgroup_release_agent);
+static void check_for_release(struct cgroup *cont);
+
/* Link structure for associating css_set objects with cgroups */
struct cg_cgroup_link {
/*
@@ -188,11 +224,8 @@ static int use_task_css_set_links;
/*
* unlink a css_set from the list and free it
*/
-static void release_css_set(struct kref *k)
+static void unlink_css_set(struct css_set *cg)
{
- struct css_set *cg = container_of(k, struct css_set, ref);
- int i;
-
write_lock(&css_set_lock);
list_del(&cg->list);
css_set_count--;
@@ -205,11 +238,39 @@ static void release_css_set(struct kre
kfree(link);
}
write_unlock(&css_set_lock);
- for (i = 0; i < CGROUP_SUBSYS_COUNT; i++)
- atomic_dec(&cg->subsys[i]->cgroup->count);
+}
+
+static void __release_css_set(struct kref *k, int taskexit)
+{
+ int i;
+ struct css_set *cg = container_of(k, struct css_set, ref);
+
+ unlink_css_set(cg);
+
+ rcu_read_lock();
+ for (i = 0; i < CGROUP_SUBSYS_COUNT; i++) {
+ struct cgroup *cont = cg->subsys[i]->cgroup;
+ if (atomic_dec_and_test(&cont->count) &&
+ notify_on_release(cont)) {
+ if (taskexit)
+ set_bit(CONT_RELEASABLE, &cont->flags);
+ check_for_release(cont);
+ }
+ }
+ rcu_read_unlock();
kfree(cg);
}
+static void release_css_set(struct kref *k)
+{
+ __release_css_set(k, 0);
+}
+
+static void release_css_set_taskexit(struct kref *k)
+{
+ __release_css_set(k, 1);
+}
+
/*
* refcounted get/put for css_set objects
*/
@@ -223,6 +284,11 @@ static inline void put_css_set(struct
kref_put(&cg->ref, release_css_set);
}
+static inline void put_css_set_taskexit(struct css_set *cg)
+{
+ kref_put(&cg->ref, release_css_set_taskexit);
+}
+
/*
* find_existing_css_set() is a helper for
* find_css_set(), and checks to see whether an existing
@@ -464,8 +530,6 @@ static struct css_set *find_css_set(
* update of a tasks cgroup pointer by attach_task()
*/
-static DEFINE_MUTEX(cgroup_mutex);
-
/**
* cgroup_lock - lock out any changes to cgroup structures
*
@@ -524,6 +588,13 @@ static void cgroup_diput(struct dentr
if (S_ISDIR(inode->i_mode)) {
struct cgroup *cont = dentry->d_fsdata;
BUG_ON(!(cgroup_is_removed(cont)));
+ /* It's possible for external users to be holding css
+ * reference counts on a cgroup; css_put() needs to
+ * be able to access the cgroup after decrementing
+ * the reference count in order to know if it needs to
+ * queue the cgroup to be handled by the release
+ * agent */
+ synchronize_rcu();
kfree(cont);
}
iput(inode);
@@ -668,6 +739,8 @@ static int cgroup_show_options(struct
seq_printf(seq, ",%s", ss->name);
if (test_bit(ROOT_NOPREFIX, &root->flags))
seq_puts(seq, ",noprefix");
+ if (strlen(root->release_agent_path))
+ seq_printf(seq, ",release_agent=%s", root->release_agent_path);
mutex_unlock(&cgroup_mutex);
return 0;
}
@@ -675,6 +748,7 @@ static int cgroup_show_options(struct
struct cgroup_sb_opts {
unsigned long subsys_bits;
unsigned long flags;
+ char *release_agent;
};
/* Convert a hierarchy specifier into a bitmask of subsystems and
@@ -686,6 +760,7 @@ static int parse_cgroupfs_options(cha
opts->subsys_bits = 0;
opts->flags = 0;
+ opts->release_agent = NULL;
while ((token = strsep(&o, ",")) != NULL) {
if (!*token)
@@ -694,6 +769,15 @@ static int parse_cgroupfs_options(cha
opts->subsys_bits = (1 << CGROUP_SUBSYS_COUNT) - 1;
} else if (!strcmp(token, "noprefix")) {
set_bit(ROOT_NOPREFIX, &opts->flags);
+ } else if (!strncmp(token, "release_agent=", 14)) {
+ /* Specifying two release agents is forbidden */
+ if (opts->release_agent)
+ return -EINVAL;
+ opts->release_agent = kzalloc(PATH_MAX, GFP_KERNEL);
+ if (!opts->release_agent)
+ return -ENOMEM;
+ strncpy(opts->release_agent, token + 14, PATH_MAX - 1);
+ opts->release_agent[PATH_MAX - 1] = 0;
} else {
struct cgroup_subsys *ss;
int i;
@@ -743,7 +827,11 @@ static int cgroup_remount(struct supe
if (!ret)
cgroup_populate_dir(cont);
+ if (opts.release_agent)
+ strcpy(root->release_agent_path, opts.release_agent);
out_unlock:
+ if (opts.release_agent)
+ kfree(opts.release_agent);
mutex_unlock(&cgroup_mutex);
mutex_unlock(&cont->dentry->d_inode->i_mutex);
return ret;
@@ -767,6 +855,7 @@ static void init_cgroup_root(struct c
INIT_LIST_HEAD(&cont->sibling);
INIT_LIST_HEAD(&cont->children);
INIT_LIST_HEAD(&cont->css_sets);
+ INIT_LIST_HEAD(&cont->release_list);
}
static int cgroup_test_super(struct super_block *sb, void *data)
@@ -841,8 +930,11 @@ static int cgroup_get_sb(struct file_
/* First find the desired set of subsystems */
ret = parse_cgroupfs_options(data, &opts);
- if (ret)
+ if (ret) {
+ if (opts.release_agent)
+ kfree(opts.release_agent);
return ret;
+ }
root = kzalloc(sizeof(*root), GFP_KERNEL);
if (!root)
@@ -851,6 +943,10 @@ static int cgroup_get_sb(struct file_
init_cgroup_root(roo
...
|
|
|
|
| [PATCH 13/29] task containersv11 simple task container debug info subsystem [message #20084 is a reply to message #20064] |
Tue, 11 September 2007 19:52 |
Paul Menage
Messages: 642
Registered: September 2006
|
Senior Member |
|
|
From: Paul Menage <menage@google.com>
This example subsystem exports debugging information as an aid to diagnosing
refcount leaks, etc, in the cgroup framework.
Signed-off-by: Paul Menage <menage@google.com>
Cc: Serge E. Hallyn <serue@us.ibm.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Dave Hansen <haveblue@us.ibm.com>
Cc: Balbir Singh <balbir@in.ibm.com>
Cc: Paul Jackson <pj@sgi.com>
Cc: Kirill Korotaev <dev@openvz.org>
Cc: Herbert Poetzl <herbert@13thfloor.at>
Cc: Srivatsa Vaddagiri <vatsa@in.ibm.com>
Cc: Cedric Le Goater <clg@fr.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---
include/linux/cgroup_subsys.h | 4 +
init/Kconfig | 10 ++
kernel/Makefile | 1
kernel/cgroup_debug.c | 97 +++++++++++++++++++++++++++++
4 files changed, 112 insertions(+)
diff -puN include/linux/cgroup_subsys.h~task-cgroupsv11-simple-task-cgroup-debug-info-subsystem include/linux/cgroup_subsys.h
--- a/include/linux/cgroup_subsys.h~task-cgroupsv11-simple-task-cgroup-debug-info-subsystem
+++ a/include/linux/cgroup_subsys.h
@@ -19,4 +19,8 @@ SUBSYS(cpuacct)
/* */
+#ifdef CONFIG_CGROUP_DEBUG
+SUBSYS(debug)
+#endif
+
/* */
diff -puN init/Kconfig~task-cgroupsv11-simple-task-cgroup-debug-info-subsystem init/Kconfig
--- a/init/Kconfig~task-cgroupsv11-simple-task-cgroup-debug-info-subsystem
+++ a/init/Kconfig
@@ -282,6 +282,16 @@ config CGROUPS
Say N if unsure.
+config CGROUP_DEBUG
+ bool "Example debug cgroup subsystem"
+ depends on CGROUPS
+ help
+ This option enables a simple cgroup subsystem that
+ exports useful debugging information about the cgroups
+ framework
+
+ Say N if unsure
+
config CPUSETS
bool "Cpuset support"
depends on SMP && CGROUPS
diff -puN kernel/Makefile~task-cgroupsv11-simple-task-cgroup-debug-info-subsystem kernel/Makefile
--- a/kernel/Makefile~task-cgroupsv11-simple-task-cgroup-debug-info-subsystem
+++ a/kernel/Makefile
@@ -39,6 +39,7 @@ obj-$(CONFIG_BSD_PROCESS_ACCT) += acct.o
obj-$(CONFIG_KEXEC) += kexec.o
obj-$(CONFIG_COMPAT) += compat.o
obj-$(CONFIG_CGROUPS) += cgroup.o
+obj-$(CONFIG_CGROUP_DEBUG) += cgroup_debug.o
obj-$(CONFIG_CPUSETS) += cpuset.o
obj-$(CONFIG_CGROUP_CPUACCT) += cpu_acct.o
obj-$(CONFIG_IKCONFIG) += configs.o
diff -puN /dev/null kernel/cgroup_debug.c
--- /dev/null
+++ a/kernel/cgroup_debug.c
@@ -0,0 +1,97 @@
+/*
+ * kernel/ccontainer_debug.c - Example cgroup subsystem that
+ * exposes debug info
+ *
+ * Copyright (C) Google Inc, 2007
+ *
+ * Developed by Paul Menage (menage@google.com)
+ *
+ */
+
+#include <linux/cgroup.h>
+#include <linux/fs.h>
+#include <linux/slab.h>
+#include <linux/rcupdate.h>
+
+#include <asm/atomic.h>
+
+static struct cgroup_subsys_state *debug_create(struct cgroup_subsys *ss,
+ struct cgroup *cont)
+{
+ struct cgroup_subsys_state *css = kzalloc(sizeof(*css), GFP_KERNEL);
+
+ if (!css)
+ return ERR_PTR(-ENOMEM);
+
+ return css;
+}
+
+static void debug_destroy(struct cgroup_subsys *ss, struct cgroup *cont)
+{
+ kfree(cont->subsys[debug_subsys_id]);
+}
+
+static u64 cgroup_refcount_read(struct cgroup *cont, struct cftype *cft)
+{
+ return atomic_read(&cont->count);
+}
+
+static u64 taskcount_read(struct cgroup *cont, struct cftype *cft)
+{
+ u64 count;
+
+ cgroup_lock();
+ count = cgroup_task_count(cont);
+ cgroup_unlock();
+ return count;
+}
+
+static u64 current_css_set_read(struct cgroup *cont, struct cftype *cft)
+{
+ return (u64)(long)current->cgroups;
+}
+
+static u64 current_css_set_refcount_read(struct cgroup *cont,
+ struct cftype *cft)
+{
+ u64 count;
+
+ rcu_read_lock();
+ count = atomic_read(¤t->cgroups->ref.refcount);
+ rcu_read_unlock();
+ return count;
+}
+
+static struct cftype files[] = {
+ {
+ .name = "cgroup_refcount",
+ .read_uint = cgroup_refcount_read,
+ },
+ {
+ .name = "taskcount",
+ .read_uint = taskcount_read,
+ },
+
+ {
+ .name = "current_css_set",
+ .read_uint = current_css_set_read,
+ },
+
+ {
+ .name = "current_css_set_refcount",
+ .read_uint = current_css_set_refcount_read,
+ },
+};
+
+static int debug_populate(struct cgroup_subsys *ss, struct cgroup *cont)
+{
+ return cgroup_add_files(cont, ss, files, ARRAY_SIZE(files));
+}
+
+struct cgroup_subsys debug_subsys = {
+ .name = "debug",
+ .create = debug_create,
+ .destroy = debug_destroy,
+ .populate = debug_populate,
+ .subsys_id = debug_subsys_id,
+};
_
--
_______________________________________________
Containers mailing list
Containers@lists.linux-foundation.org
https://lists.linux-foundation.org/mailman/listinfo/containers
|
|
|
|
| [PATCH 06/29] task containersv11 add procfs interface [message #20085 is a reply to message #20064] |
Tue, 11 September 2007 19:52 |
Paul Menage
Messages: 642
Registered: September 2006
|
Senior Member |
|
|
From: Paul Menage <menage@google.com>
Add:
/proc/cgroups - general system info
/proc/*/cgroup - per-task cgroup membership info
Signed-off-by: Paul Menage <menage@google.com>
Cc: Serge E. Hallyn <serue@us.ibm.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Dave Hansen <haveblue@us.ibm.com>
Cc: Balbir Singh <balbir@in.ibm.com>
Cc: Paul Jackson <pj@sgi.com>
Cc: Kirill Korotaev <dev@openvz.org>
Cc: Herbert Poetzl <herbert@13thfloor.at>
Cc: Srivatsa Vaddagiri <vatsa@in.ibm.com>
Cc: Cedric Le Goater <clg@fr.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---
fs/proc/base.c | 7 +
include/linux/cgroup.h | 2
kernel/cgroup.c | 132 ++++++++++++++++++++++++++++++++++++
3 files changed, 141 insertions(+)
diff -puN fs/proc/base.c~task-cgroupsv11-add-procfs-interface fs/proc/base.c
--- a/fs/proc/base.c~task-cgroupsv11-add-procfs-interface
+++ a/fs/proc/base.c
@@ -67,6 +67,7 @@
#include <linux/mount.h>
#include <linux/security.h>
#include <linux/ptrace.h>
+#include <linux/cgroup.h>
#include <linux/cpuset.h>
#include <linux/audit.h>
#include <linux/poll.h>
@@ -2051,6 +2052,9 @@ static const struct pid_entry tgid_base_
#ifdef CONFIG_CPUSETS
REG("cpuset", S_IRUGO, cpuset),
#endif
+#ifdef CONFIG_CGROUPS
+ REG("cgroup", S_IRUGO, cgroup),
+#endif
INF("oom_score", S_IRUGO, oom_score),
REG("oom_adj", S_IRUGO|S_IWUSR, oom_adjust),
#ifdef CONFIG_AUDITSYSCALL
@@ -2340,6 +2344,9 @@ static const struct pid_entry tid_base_s
#ifdef CONFIG_CPUSETS
REG("cpuset", S_IRUGO, cpuset),
#endif
+#ifdef CONFIG_CGROUPS
+ REG("cgroup", S_IRUGO, cgroup),
+#endif
INF("oom_score", S_IRUGO, oom_score),
REG("oom_adj", S_IRUGO|S_IWUSR, oom_adjust),
#ifdef CONFIG_AUDITSYSCALL
diff -puN include/linux/cgroup.h~task-cgroupsv11-add-procfs-interface include/linux/cgroup.h
--- a/include/linux/cgroup.h~task-cgroupsv11-add-procfs-interface
+++ a/include/linux/cgroup.h
@@ -29,6 +29,8 @@ extern void cgroup_fork(struct task_s
extern void cgroup_fork_callbacks(struct task_struct *p);
extern void cgroup_exit(struct task_struct *p, int run_callbacks);
+extern struct file_operations proc_cgroup_operations;
+
/* Per-subsystem/per-cgroup state maintained by the system. */
struct cgroup_subsys_state {
/* The cgroup that this subsystem is attached to. Useful
diff -puN kernel/cgroup.c~task-cgroupsv11-add-procfs-interface kernel/cgroup.c
--- a/kernel/cgroup.c~task-cgroupsv11-add-procfs-interface
+++ a/kernel/cgroup.c
@@ -33,6 +33,7 @@
#include <linux/mutex.h>
#include <linux/mount.h>
#include <linux/pagemap.h>
+#include <linux/proc_fs.h>
#include <linux/rcupdate.h>
#include <linux/sched.h>
#include <linux/seq_file.h>
@@ -247,6 +248,7 @@ static int cgroup_mkdir(struct inode
static int cgroup_rmdir(struct inode *unused_dir, struct dentry *dentry);
static int cgroup_populate_dir(struct cgroup *cont);
static struct inode_operations cgroup_dir_inode_operations;
+static struct file_operations proc_cgroupstats_operations;
static struct inode *cgroup_new_inode(mode_t mode, struct super_block *sb)
{
@@ -1576,6 +1578,7 @@ int __init cgroup_init(void)
{
int err;
int i;
+ struct proc_dir_entry *entry;
for (i = 0; i < CGROUP_SUBSYS_COUNT; i++) {
struct cgroup_subsys *ss = subsys[i];
@@ -1587,10 +1590,139 @@ int __init cgroup_init(void)
if (err < 0)
goto out;
+ entry = create_proc_entry("cgroups", 0, NULL);
+ if (entry)
+ entry->proc_fops = &proc_cgroupstats_operations;
+
out:
return err;
}
+/*
+ * proc_cgroup_show()
+ * - Print task's cgroup paths into seq_file, one line for each hierarchy
+ * - Used for /proc/<pid>/cgroup.
+ * - No need to task_lock(tsk) on this tsk->cgroup reference, as it
+ * doesn't really matter if tsk->cgroup changes after we read it,
+ * and we take cgroup_mutex, keeping attach_task() from changing it
+ * anyway. No need to check that tsk->cgroup != NULL, thanks to
+ * the_top_cgroup_hack in cgroup_exit(), which sets an exiting tasks
+ * cgroup to top_cgroup.
+ */
+
+/* TODO: Use a proper seq_file iterator */
+static int proc_cgroup_show(struct seq_file *m, void *v)
+{
+ struct pid *pid;
+ struct task_struct *tsk;
+ char *buf;
+ int retval;
+ struct cgroupfs_root *root;
+
+ retval = -ENOMEM;
+ buf = kmalloc(PAGE_SIZE, GFP_KERNEL);
+ if (!buf)
+ goto out;
+
+ retval = -ESRCH;
+ pid = m->private;
+ tsk = get_pid_task(pid, PIDTYPE_PID);
+ if (!tsk)
+ goto out_free;
+
+ retval = 0;
+
+ mutex_lock(&cgroup_mutex);
+
+ for_each_root(root) {
+ struct cgroup_subsys *ss;
+ struct cgroup *cont;
+ int subsys_id;
+ int count = 0;
+
+ /* Skip this hierarchy if it has no active subsystems */
+ if (!root->actual_subsys_bits)
+ continue;
+ for_each_subsys(root, ss)
+ seq_printf(m, "%s%s", count++ ? "," : "", ss->name);
+ seq_putc(m, ':');
+ get_first_subsys(&root->top_cgroup, NULL, &subsys_id);
+ cont = task_cgroup(tsk, subsys_id);
+ retval = cgroup_path(cont, buf, PAGE_SIZE);
+ if (retval < 0)
+ goto out_unlock;
+ seq_puts(m, buf);
+ seq_putc(m, '\n');
+ }
+
+out_unlock:
+ mutex_unlock(&cgroup_mutex);
+ put_task_struct(tsk);
+out_free:
+ kfree(buf);
+out:
+ return retval;
+}
+
+static int cgroup_open(struct inode *inode, struct file *file)
+{
+ struct pid *pid = PROC_I(inode)->pid;
+ return single_open(file, proc_cgroup_show, pid);
+}
+
+struct file_operations proc_cgroup_operations = {
+ .open = cgroup_open,
+ .read = seq_read,
+ .llseek = seq_lseek,
+ .release = single_release,
+};
+
+/* Display information about each subsystem and each hierarchy */
+static int proc_cgroupstats_show(struct seq_file *m, void *v)
+{
+ int i;
+ struct cgroupfs_root *root;
+
+ mutex_lock(&cgroup_mutex);
+ seq_puts(m, "Hierarchies:\n");
+ for_each_root(root) {
+ struct cgroup_subsys *ss;
+ int first = 1;
+ seq_printf(m, "%p: bits=%lx cgroups=%d (", root,
+ root->subsys_bits, root->number_of_cgroups);
+ for_each_subsys(root, ss) {
+ seq_printf(m, "%s%s", first ? "" : ", ", ss->name);
+ first = false;
+ }
+ seq_putc(m, ')');
+ if (root->sb) {
+ seq_printf(m, " s_active=%d",
+ atomic_read(&root->sb->s_active));
+ }
+ seq_putc(m, '\n');
+ }
+ seq_puts(m, "Subsystems:\n");
+ for (i = 0; i < CGROUP_SUBSYS_COUNT; i++) {
+ struct cgroup_subsys *ss = subsys[i];
+ seq_printf(m, "%d: name=%s hierarchy=%p\n",
+ i, ss->name, ss->root);
+ }
+ mutex_unlock(&cgroup_mutex);
+ return 0;
+}
+
+static int cgroupstats_open(struct inode *inode, struct file *file)
+{
+ return single_open(file, proc_cgroupstats_show, 0);
+}
+
+static struct file_operations proc_cgroupstats_operations = {
+ .open = cgroupstats_open,
+ .read = seq_read,
+ .llseek = seq_lseek,
+ .release = single_release,
+};
+
/**
* cgroup_fork - attach newly forked task to its parents cgroup.
* @tsk: pointer to task_struct of forking parent process.
_
--
_______________________________________________
Containers mailing list
Containers@lists.linux-foundation.org
https://lists.linux-foundation.org/mailman/listinfo/containers
|
|
|
|
| [PATCH 22/29] memory controller accounting setup v7 [message #20086 is a reply to message #20064] |
Tue, 11 September 2007 19:53 |
Paul Menage
Messages: 642
Registered: September 2006
|
Senior Member |
|
|
From: Pavel Emelianov <xemul@openvz.org>
Basic setup routines, the mm_struct has a pointer to the cgroup that
it belongs to and the the page has a page_cgroup associated with it.
Signed-off-by: Pavel Emelianov <xemul@openvz.org>
Signed-off-by: Balbir Singh <balbir@linux.vnet.ibm.com>
Cc: Paul Menage <menage@google.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Cc: Kirill Korotaev <dev@sw.ru>
Cc: Herbert Poetzl <herbert@13thfloor.at>
Cc: David Rientjes <rientjes@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---
include/linux/memcontrol.h | 36 ++++++++++++++++++++++
include/linux/mm_types.h | 6 +++
include/linux/sched.h | 1
kernel/fork.c | 11 ++++--
mm/memcontrol.c | 57 ++++++++++++++++++++++++++++++++---
5 files changed, 104 insertions(+), 7 deletions(-)
diff -puN include/linux/memcontrol.h~memory-controller-accounting-setup-v7 include/linux/memcontrol.h
--- a/include/linux/memcontrol.h~memory-controller-accounting-setup-v7
+++ a/include/linux/memcontrol.h
@@ -3,6 +3,9 @@
* Copyright IBM Corporation, 2007
* Author Balbir Singh <balbir@linux.vnet.ibm.com>
*
+ * Copyright 2007 OpenVZ SWsoft Inc
+ * Author: Pavel Emelianov <xemul@openvz.org>
+ *
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2 of the License, or
@@ -17,5 +20,38 @@
#ifndef _LINUX_MEMCONTROL_H
#define _LINUX_MEMCONTROL_H
+struct mem_cgroup;
+struct page_cgroup;
+
+#ifdef CONFIG_CGROUP_MEM_CONT
+
+extern void mm_init_cgroup(struct mm_struct *mm, struct task_struct *p);
+extern void mm_free_cgroup(struct mm_struct *mm);
+extern void page_assign_page_cgroup(struct page *page,
+ struct page_cgroup *pc);
+extern struct page_cgroup *page_get_page_cgroup(struct page *page);
+
+#else /* CONFIG_CGROUP_MEM_CONT */
+static inline void mm_init_cgroup(struct mm_struct *mm,
+ struct task_struct *p)
+{
+}
+
+static inline void mm_free_cgroup(struct mm_struct *mm)
+{
+}
+
+static inline void page_assign_page_cgroup(struct page *page,
+ struct page_cgroup *pc)
+{
+}
+
+static inline struct page_cgroup *page_get_page_cgroup(struct page *page)
+{
+ return NULL;
+}
+
+#endif /* CONFIG_CGROUP_MEM_CONT */
+
#endif /* _LINUX_MEMCONTROL_H */
diff -puN include/linux/mm_types.h~memory-controller-accounting-setup-v7 include/linux/mm_types.h
--- a/include/linux/mm_types.h~memory-controller-accounting-setup-v7
+++ a/include/linux/mm_types.h
@@ -83,6 +83,9 @@ struct page {
void *virtual; /* Kernel virtual address (NULL if
not kmapped, ie. highmem) */
#endif /* WANT_PAGE_VIRTUAL */
+#ifdef CONFIG_CGROUP_MEM_CONT
+ unsigned long page_cgroup;
+#endif
};
/*
@@ -214,6 +217,9 @@ struct mm_struct {
/* aio bits */
rwlock_t ioctx_list_lock;
struct kioctx *ioctx_list;
+#ifdef CONFIG_CGROUP_MEM_CONT
+ struct mem_cgroup *mem_cgroup;
+#endif
};
#endif /* _LINUX_MM_TYPES_H */
diff -puN include/linux/sched.h~memory-controller-accounting-setup-v7 include/linux/sched.h
--- a/include/linux/sched.h~memory-controller-accounting-setup-v7
+++ a/include/linux/sched.h
@@ -88,6 +88,7 @@ struct sched_param {
#include <asm/processor.h>
+struct mem_cgroup;
struct exec_domain;
struct futex_pi_state;
struct bio;
diff -puN kernel/fork.c~memory-controller-accounting-setup-v7 kernel/fork.c
--- a/kernel/fork.c~memory-controller-accounting-setup-v7
+++ a/kernel/fork.c
@@ -51,6 +51,7 @@
#include <linux/random.h>
#include <linux/tty.h>
#include <linux/proc_fs.h>
+#include <linux/memcontrol.h>
#include <asm/pgtable.h>
#include <asm/pgalloc.h>
@@ -329,7 +330,7 @@ __cacheline_aligned_in_smp DEFINE_SPINLO
#include <linux/init_task.h>
-static struct mm_struct * mm_init(struct mm_struct * mm)
+static struct mm_struct * mm_init(struct mm_struct * mm, struct task_struct *p)
{
atomic_set(&mm->mm_users, 1);
atomic_set(&mm->mm_count, 1);
@@ -346,11 +347,14 @@ static struct mm_struct * mm_init(struct
mm->ioctx_list = NULL;
mm->free_area_cache = TASK_UNMAPPED_BASE;
mm->cached_hole_size = ~0UL;
+ mm_init_cgroup(mm, p);
if (likely(!mm_alloc_pgd(mm))) {
mm->def_flags = 0;
return mm;
}
+
+ mm_free_cgroup(mm);
free_mm(mm);
return NULL;
}
@@ -365,7 +369,7 @@ struct mm_struct * mm_alloc(void)
mm = allocate_mm();
if (mm) {
memset(mm, 0, sizeof(*mm));
- mm = mm_init(mm);
+ mm = mm_init(mm, current);
}
return mm;
}
@@ -379,6 +383,7 @@ void fastcall __mmdrop(struct mm_struct
{
BUG_ON(mm == &init_mm);
mm_free_pgd(mm);
+ mm_free_cgroup(mm);
destroy_context(mm);
free_mm(mm);
}
@@ -499,7 +504,7 @@ static struct mm_struct *dup_mm(struct t
mm->token_priority = 0;
mm->last_interval = 0;
- if (!mm_init(mm))
+ if (!mm_init(mm, tsk))
goto fail_nomem;
if (init_new_context(tsk, mm))
diff -puN mm/memcontrol.c~memory-controller-accounting-setup-v7 mm/memcontrol.c
--- a/mm/memcontrol.c~memory-controller-accounting-setup-v7
+++ a/mm/memcontrol.c
@@ -3,6 +3,9 @@
* Copyright IBM Corporation, 2007
* Author Balbir Singh <balbir@linux.vnet.ibm.com>
*
+ * Copyright 2007 OpenVZ SWsoft Inc
+ * Author: Pavel Emelianov <xemul@openvz.org>
+ *
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2 of the License, or
@@ -17,6 +20,7 @@
#include <linux/res_counter.h>
#include <linux/memcontrol.h>
#include <linux/cgroup.h>
+#include <linux/mm.h>
struct cgroup_subsys mem_cgroup_subsys;
@@ -35,6 +39,13 @@ struct mem_cgroup {
* the counter to account for memory usage
*/
struct res_counter res;
+ /*
+ * Per cgroup active and inactive list, similar to the
+ * per zone LRU lists.
+ * TODO: Consider making these lists per zone
+ */
+ struct list_head active_list;
+ struct list_head inactive_list;
};
/*
@@ -56,6 +67,37 @@ struct mem_cgroup *mem_cgroup_from
css);
}
+static inline
+struct mem_cgroup *mem_cgroup_from_task(struct task_struct *p)
+{
+ return container_of(task_subsys_state(p, mem_cgroup_subsys_id),
+ struct mem_cgroup, css);
+}
+
+void mm_init_cgroup(struct mm_struct *mm, struct task_struct *p)
+{
+ struct mem_cgroup *mem;
+
+ mem = mem_cgroup_from_task(p);
+ css_get(&mem->css);
+ mm->mem_cgroup = mem;
+}
+
+void mm_free_cgroup(struct mm_struct *mm)
+{
+ css_put(&mm->mem_cgroup->css);
+}
+
+void page_assign_page_cgroup(struct page *page, struct page_cgroup *pc)
+{
+ page->page_cgroup = (unsigned long)pc;
+}
+
+struct page_cgroup *page_get_page_cgroup(struct page *page)
+{
+ return page->page_cgroup;
+}
+
static ssize_t mem_cgroup_read(struct cgroup *cont, struct cftype *cft,
struct file *file, char __user *userbuf, size_t nbytes,
loff_t *ppos)
@@ -91,14 +133,21 @@ static struct cftype mem_cgroup_files
},
};
+static struct mem_cgroup init_mem_cgroup;
+
static struct cgroup_subsys_state *
mem_cgroup_create(struct cgroup_subsys *ss, struct cgroup *cont)
{
struct mem_cgroup *mem;
- mem = kzalloc(sizeof(struct mem_cgroup), GFP_KERNEL);
- if (!mem)
- return -ENOMEM;
+ if (unlikely((cont->parent) == NULL)) {
+ mem = &init_mem_cgroup;
+ init_mm.mem_cgroup = mem;
+ } else
+ mem = kzalloc(sizeof(struct mem_cgroup), GFP_KERNEL);
+
+ if (mem == NULL)
+ return NULL;
res_counter_init(&mem->res);
return &mem->css;
@@ -123,5 +172,5 @@ struct cgroup_subsys mem_cgroup_su
.create = mem_cgroup_create,
.destroy = mem_cgroup_destroy,
.populate = mem_cgroup_populate,
- .early_init = 0,
+ .early_init = 1,
};
_
--
_______________________________________________
Containers mailing list
Containers@lists.linux-foundation.org
https://lists.linux-foundation.org/mailman/listinfo/containers
|
|
|
|
| [PATCH 27/29] memory controller oom handling v7 [message #20087 is a reply to message #20064] |
Tue, 11 September 2007 19:53 |
Paul Menage
Messages: 642
Registered: September 2006
|
Senior Member |
|
|
From: Pavel Emelianov <xemul@openvz.org>
Out of memory handling for cgroups over their limit. A task from the
cgroup over limit is chosen using the existing OOM logic and killed.
TODO:
1. As discussed in the OLS BOF session, consider implementing a user
space policy for OOM handling.
Signed-off-by: Pavel Emelianov <xemul@openvz.org>
Signed-off-by: Balbir Singh <balbir@linux.vnet.ibm.com>
Cc: Paul Menage <menage@google.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Cc: Kirill Korotaev <dev@sw.ru>
Cc: Herbert Poetzl <herbert@13thfloor.at>
Cc: David Rientjes <rientjes@google.com>
Cc: Vaidyanathan Srinivasan <svaidy@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---
include/linux/memcontrol.h | 1
mm/memcontrol.c | 1
mm/oom_kill.c | 42 +++++++++++++++++++++++++++++++----
3 files changed, 40 insertions(+), 4 deletions(-)
diff -puN include/linux/memcontrol.h~memory-controller-oom-handling-v7 include/linux/memcontrol.h
--- a/include/linux/memcontrol.h~memory-controller-oom-handling-v7
+++ a/include/linux/memcontrol.h
@@ -39,6 +39,7 @@ extern unsigned long mem_cgroup_isola
int mode, struct zone *z,
struct mem_cgroup *mem_cont,
int active);
+extern void mem_cgroup_out_of_memory(struct mem_cgroup *mem);
static inline void mem_cgroup_uncharge_page(struct page *page)
{
diff -puN mm/memcontrol.c~memory-controller-oom-handling-v7 mm/memcontrol.c
--- a/mm/memcontrol.c~memory-controller-oom-handling-v7
+++ a/mm/memcontrol.c
@@ -322,6 +322,7 @@ int mem_cgroup_charge(struct page *pa
}
css_put(&mem->css);
+ mem_cgroup_out_of_memory(mem);
goto free_pc;
}
diff -puN mm/oom_kill.c~memory-controller-oom-handling-v7 mm/oom_kill.c
--- a/mm/oom_kill.c~memory-controller-oom-handling-v7
+++ a/mm/oom_kill.c
@@ -25,6 +25,7 @@
#include <linux/cpuset.h>
#include <linux/module.h>
#include <linux/notifier.h>
+#include <linux/memcontrol.h>
int sysctl_panic_on_oom;
/* #define DEBUG */
@@ -48,7 +49,8 @@ int sysctl_panic_on_oom;
* of least surprise ... (be careful when you change it)
*/
-unsigned long badness(struct task_struct *p, unsigned long uptime)
+unsigned long badness(struct task_struct *p, unsigned long uptime,
+ struct mem_cgroup *mem)
{
unsigned long points, cpu_time, run_time, s;
struct mm_struct *mm;
@@ -61,6 +63,13 @@ unsigned long badness(struct task_struct
return 0;
}
+#ifdef CONFIG_CGROUP_MEM_CONT
+ if (mem != NULL && mm->mem_cgroup != mem) {
+ task_unlock(p);
+ return 0;
+ }
+#endif
+
/*
* The memory size of the process is the basis for the badness.
*/
@@ -198,7 +207,8 @@ static inline int constrained_alloc(stru
*
* (not docbooked, we don't want this one cluttering up the manual)
*/
-static struct task_struct *select_bad_process(unsigned long *ppoints)
+static struct task_struct *select_bad_process(unsigned long *ppoints,
+ struct mem_cgroup *mem)
{
struct task_struct *g, *p;
struct task_struct *chosen = NULL;
@@ -252,7 +262,7 @@ static struct task_struct *select_bad_pr
if (p->oomkilladj == OOM_DISABLE)
continue;
- points = badness(p, uptime.tv_sec);
+ points = badness(p, uptime.tv_sec, mem);
if (points > *ppoints || !chosen) {
chosen = p;
*ppoints = points;
@@ -364,6 +374,30 @@ static int oom_kill_process(struct task_
return oom_kill_task(p);
}
+#ifdef CONFIG_CGROUP_MEM_CONT
+void mem_cgroup_out_of_memory(struct mem_cgroup *mem)
+{
+ unsigned long points = 0;
+ struct task_struct *p;
+
+ cgroup_lock();
+ rcu_read_lock();
+retry:
+ p = select_bad_process(&points, mem);
+ if (PTR_ERR(p) == -1UL)
+ goto out;
+
+ if (!p)
+ p = current;
+
+ if (oom_kill_process(p, points, "Memory cgroup out of memory"))
+ goto retry;
+out:
+ rcu_read_unlock();
+ cgroup_unlock();
+}
+#endif
+
static BLOCKING_NOTIFIER_HEAD(oom_notify_list);
int register_oom_notifier(struct notifier_block *nb)
@@ -436,7 +470,7 @@ retry:
* Rambo mode: Shoot down a process and hope it solves whatever
* issues we may have.
*/
- p = select_bad_process(&points);
+ p = select_bad_process(&points, NULL);
if (PTR_ERR(p) == -1UL)
goto out;
_
--
_______________________________________________
Containers mailing list
Containers@lists.linux-foundation.org
https://lists.linux-foundation.org/mailman/listinfo/containers
|
|
|
|
| [PATCH 17/29] containers implement namespace tracking subsystem fix order of container subsystems in [message #20088 is a reply to message #20064] |
Tue, 11 September 2007 19:52 |
Paul Menage
Messages: 642
Registered: September 2006
|
Senior Member |
|
|
From: Cedric Le Goater <clg@fr.ibm.com>
some cosmetic changes to init/Kconfig to make the subsystem list
under cgroups look nicer when CONFIG_CGROUPS=y.
also changed the 'select' to 'depends on' for the namespace
subsystem.
Signed-off-by: Cedric Le Goater <clg@fr.ibm.com>
Acked-by: Paul Menage <menage@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---
init/Kconfig | 32 ++++++++++++++++----------------
1 file changed, 16 insertions(+), 16 deletions(-)
diff -puN init/Kconfig~cgroups-implement-namespace-tracking-subsystem-fix-order-of-cgroup-subsystems-in-init-kconfig init/Kconfig
--- a/init/Kconfig~cgroups-implement-namespace-tracking-subsystem-fix-order-of-cgroup-subsystems-in-init-kconfig
+++ a/init/Kconfig
@@ -292,6 +292,22 @@ config CGROUP_DEBUG
Say N if unsure
+config CGROUP_NS
+ bool "Namespace cgroup subsystem"
+ depends on CGROUPS
+ help
+ Provides a simple namespace cgroup subsystem to
+ provide hierarchical naming of sets of namespaces,
+ for instance virtual servers and checkpoint/restart
+ jobs.
+
+config CGROUP_CPUACCT
+ bool "Simple CPU accounting cgroup subsystem"
+ depends on CGROUPS
+ help
+ Provides a simple Resource Controller for monitoring the
+ total CPU consumed by the tasks in a cgroup
+
config CPUSETS
bool "Cpuset support"
depends on SMP && CGROUPS
@@ -323,27 +339,11 @@ 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 CGROUP_NS
- bool "Namespace cgroup subsystem"
- select CGROUPS
- help
- Provides a simple namespace cgroup subsystem to
- provide hierarchical naming of sets of namespaces,
- for instance virtual servers and checkpoint/restart
- jobs.
-
config PROC_PID_CPUSET
bool "Include legacy /proc/<pid>/cpuset file"
depends on CPUSETS
default y
-config CGROUP_CPUACCT
- bool "Simple CPU accounting cgroup subsystem"
- depends on CGROUPS
- help
- Provides a simple Resource Controller for monitoring the
- total CPU consumed by the tasks in a cgroup
-
config RELAY
bool "Kernel->user space relay support (formerly relayfs)"
help
_
--
_______________________________________________
Containers mailing list
Containers@lists.linux-foundation.org
https://lists.linux-foundation.org/mailman/listinfo/containers
|
|
|
|
| [PATCH 28/29] memory controller add switch to control what type of pages to limit v7 [message #20089 is a reply to message #20064] |
Tue, 11 September 2007 19:53 |
Paul Menage
Messages: 642
Registered: September 2006
|
Senior Member |
|
|
From: Balbir Singh <balbir@linux.vnet.ibm.com>
Choose if we want cached pages to be accounted or not. By default both are
accounted for. A new set of tunables are added.
echo -n 1 > mem_control_type
switches the accounting to account for only mapped pages
echo -n 3 > mem_control_type
switches the behaviour back
Signed-off-by: Balbir Singh <balbir@linux.vnet.ibm.com>
Cc: Pavel Emelianov <xemul@openvz.org>
Cc: Paul Menage <menage@google.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Cc: Kirill Korotaev <dev@sw.ru>
Cc: Herbert Poetzl <herbert@13thfloor.at>
Cc: David Rientjes <rientjes@google.com>
Cc: Vaidyanathan Srinivasan <svaidy@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---
include/linux/memcontrol.h | 9 +++
mm/filemap.c | 2
mm/memcontrol.c | 92 +++++++++++++++++++++++++++++++++++
mm/swap_state.c | 2
4 files changed, 103 insertions(+), 2 deletions(-)
diff -puN include/linux/memcontrol.h~memory-controller-add-switch-to-control-what-type-of-pages-to-limit-v7 include/linux/memcontrol.h
--- a/include/linux/memcontrol.h~memory-controller-add-switch-to-control-what-type-of-pages-to-limit-v7
+++ a/include/linux/memcontrol.h
@@ -20,6 +20,8 @@
#ifndef _LINUX_MEMCONTROL_H
#define _LINUX_MEMCONTROL_H
+#include <linux/mm.h>
+
struct mem_cgroup;
struct page_cgroup;
@@ -40,6 +42,7 @@ extern unsigned long mem_cgroup_isola
struct mem_cgroup *mem_cont,
int active);
extern void mem_cgroup_out_of_memory(struct mem_cgroup *mem);
+extern int mem_cgroup_cache_charge(struct page *page, struct mm_struct *mm);
static inline void mem_cgroup_uncharge_page(struct page *page)
{
@@ -84,6 +87,12 @@ static inline void mem_cgroup_move_li
{
}
+static inline int mem_cgroup_cache_charge(struct page *page,
+ struct mm_struct *mm)
+{
+ return 0;
+}
+
#endif /* CONFIG_CGROUP_MEM_CONT */
#endif /* _LINUX_MEMCONTROL_H */
diff -puN mm/filemap.c~memory-controller-add-switch-to-control-what-type-of-pages-to-limit-v7 mm/filemap.c
--- a/mm/filemap.c~memory-controller-add-switch-to-control-what-type-of-pages-to-limit-v7
+++ a/mm/filemap.c
@@ -444,7 +444,7 @@ int add_to_page_cache(struct page *page,
if (error == 0) {
- error = mem_cgroup_charge(page, current->mm);
+ error = mem_cgroup_cache_charge(page, current->mm);
if (error)
goto out;
diff -puN mm/memcontrol.c~memory-controller-add-switch-to-control-what-type-of-pages-to-limit-v7 mm/memcontrol.c
--- a/mm/memcontrol.c~memory-controller-add-switch-to-control-what-type-of-pages-to-limit-v7
+++ a/mm/memcontrol.c
@@ -28,6 +28,8 @@
#include <linux/spinlock.h>
#include <linux/fs.h>
+#include <asm/uaccess.h>
+
struct cgroup_subsys mem_cgroup_subsys;
static const int MEM_CGROUP_RECLAIM_RETRIES = 5;
@@ -59,6 +61,7 @@ struct mem_cgroup {
* spin_lock to protect the per cgroup LRU
*/
spinlock_t lru_lock;
+ unsigned long control_type; /* control RSS or RSS+Pagecache */
};
/*
@@ -81,6 +84,15 @@ struct page_cgroup {
/* mapped and cached states */
};
+enum {
+ MEM_CGROUP_TYPE_UNSPEC = 0,
+ MEM_CGROUP_TYPE_MAPPED,
+ MEM_CGROUP_TYPE_CACHED,
+ MEM_CGROUP_TYPE_ALL,
+ MEM_CGROUP_TYPE_MAX,
+} mem_control_type;
+
+static struct mem_cgroup init_mem_cgroup;
static inline
struct mem_cgroup *mem_cgroup_from_cont(struct cgroup *cont)
@@ -361,6 +373,22 @@ err:
}
/*
+ * See if the cached pages should be charged at all?
+ */
+int mem_cgroup_cache_charge(struct page *page, struct mm_struct *mm)
+{
+ struct mem_cgroup *mem;
+ if (!mm)
+ mm = &init_mm;
+
+ mem = rcu_dereference(mm->mem_cgroup);
+ if (mem->control_type == MEM_CGROUP_TYPE_ALL)
+ return mem_cgroup_charge(page, mm);
+ else
+ return 0;
+}
+
+/*
* Uncharging is always a welcome operation, we never complain, simply
* uncharge.
*/
@@ -370,6 +398,10 @@ void mem_cgroup_uncharge(struct page_
struct page *page;
unsigned long flags;
+ /*
+ * This can handle cases when a page is not charged at all and we
+ * are switching between handling the control_type.
+ */
if (!pc)
return;
@@ -405,6 +437,60 @@ static ssize_t mem_cgroup_write(struc
cft->private, userbuf, nbytes, ppos);
}
+static ssize_t mem_control_type_write(struct cgroup *cont,
+ struct cftype *cft, struct file *file,
+ const char __user *userbuf,
+ size_t nbytes, loff_t *pos)
+{
+ int ret;
+ char *buf, *end;
+ unsigned long tmp;
+ struct mem_cgroup *mem;
+
+ mem = mem_cgroup_from_cont(cont);
+ buf = kmalloc(nbytes + 1, GFP_KERNEL);
+ ret = -ENOMEM;
+ if (buf == NULL)
+ goto out;
+
+ buf[nbytes] = 0;
+ ret = -EFAULT;
+ if (copy_from_user(buf, userbuf, nbytes))
+ goto out_free;
+
+ ret = -EINVAL;
+ tmp = simple_strtoul(buf, &end, 10);
+ if (*end != '\0')
+ goto out_free;
+
+ if (tmp <= MEM_CGROUP_TYPE_UNSPEC || tmp >= MEM_CGROUP_TYPE_MAX)
+ goto out_free;
+
+ mem->control_type = tmp;
+ ret = nbytes;
+out_free:
+ kfree(buf);
+out:
+ return ret;
+}
+
+static ssize_t mem_control_type_read(struct cgroup *cont,
+ struct cftype *cft,
+ struct file *file, char __user *userbuf,
+ size_t nbytes, loff_t *ppos)
+{
+ unsigned long val;
+ char buf[64], *s;
+ struct mem_cgroup *mem;
+
+ mem = mem_cgroup_from_cont(cont);
+ s = buf;
+ val = mem->control_type;
+ s += sprintf(s, "%lu\n", val);
+ return simple_read_from_buffer((void __user *)userbuf, nbytes,
+ ppos, buf, s - buf);
+}
+
static struct cftype mem_cgroup_files[] = {
{
.name = "usage",
@@ -422,6 +508,11 @@ static struct cftype mem_cgroup_files
.private = RES_FAILCNT,
.read = mem_cgroup_read,
},
+ {
+ .name = "control_type",
+ .write = mem_control_type_write,
+ .read = mem_control_type_read,
+ },
};
static struct mem_cgroup init_mem_cgroup;
@@ -444,6 +535,7 @@ mem_cgroup_create(struct cgroup_su
INIT_LIST_HEAD(&mem->active_list);
INIT_LIST_HEAD(&mem->inactive_list);
spin_lock_init(&mem->lru_lock);
+ mem->control_type = MEM_CGROUP_TYPE_ALL;
return &mem->css;
}
diff -puN mm/swap_state.c~memory-controller-add-switch-to-control-what-type-of-pages-to-limit-v7 mm/swap_state.c
--- a/mm/swap_state.c~memory-controller-add-switch-to-control-what-type-of-pages-to-limit-v7
+++ a/mm/swap_state.c
@@ -81,7 +81,7 @@ static int __add_to_swap_cache(struct pa
error = radix_tree_preload(gfp_mask);
if (!error) {
- error = mem_cgroup_charge(page, current->mm);
+ error = mem_cgroup_cache_charge(page, current->mm);
if (error)
goto out;
_
--
_______________________________________________
Containers mailing list
Containers@lists.linux-foundation.org
https://lists.linux-foundation.org/mailman/listinfo/containers
|
|
|
|
| [PATCH 12/29] task containersv11 example cpu accounting subsystem [message #20090 is a reply to message #20064] |
Tue, 11 September 2007 19:52 |
Paul Menage
Messages: 642
Registered: September 2006
|
Senior Member |
|
|
From: Paul Menage <menage@google.com>
This example demonstrates how to use the generic cgroup subsystem for a
simple resource tracker that counts, for the processes in a cgroup, the
total CPU time used and the %CPU used in the last complete 10 second interval.
Portions contributed by Balbir Singh <balbir@in.ibm.com>
Signed-off-by: Paul Menage <menage@google.com>
Cc: Serge E. Hallyn <serue@us.ibm.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Dave Hansen <haveblue@us.ibm.com>
Cc: Balbir Singh <balbir@in.ibm.com>
Cc: Paul Jackson <pj@sgi.com>
Cc: Kirill Korotaev <dev@openvz.org>
Cc: Herbert Poetzl <herbert@13thfloor.at>
Cc: Srivatsa Vaddagiri <vatsa@in.ibm.com>
Cc: Cedric Le Goater <clg@fr.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---
include/linux/cgroup_subsys.h | 6
include/linux/cpu_acct.h | 14 ++
init/Kconfig | 7 +
kernel/Makefile | 1
kernel/cpu_acct.c | 186 +++++++++++++++++++++++++++++
kernel/sched.c | 14 +-
6 files changed, 225 insertions(+), 3 deletions(-)
diff -puN include/linux/cgroup_subsys.h~task-cgroupsv11-example-cpu-accounting-subsystem include/linux/cgroup_subsys.h
--- a/include/linux/cgroup_subsys.h~task-cgroupsv11-example-cpu-accounting-subsystem
+++ a/include/linux/cgroup_subsys.h
@@ -13,4 +13,10 @@ SUBSYS(cpuset)
/* */
+#ifdef CONFIG_CGROUP_CPUACCT
+SUBSYS(cpuacct)
+#endif
+
+/* */
+
/* */
diff -puN /dev/null include/linux/cpu_acct.h
--- /dev/null
+++ a/include/linux/cpu_acct.h
@@ -0,0 +1,14 @@
+
+#ifndef _LINUX_CPU_ACCT_H
+#define _LINUX_CPU_ACCT_H
+
+#include <linux/cgroup.h>
+#include <asm/cputime.h>
+
+#ifdef CONFIG_CGROUP_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
diff -puN init/Kconfig~task-cgroupsv11-example-cpu-accounting-subsystem init/Kconfig
--- a/init/Kconfig~task-cgroupsv11-example-cpu-accounting-subsystem
+++ a/init/Kconfig
@@ -318,6 +318,13 @@ config PROC_PID_CPUSET
depends on CPUSETS
default y
+config CGROUP_CPUACCT
+ bool "Simple CPU accounting cgroup subsystem"
+ depends on CGROUPS
+ help
+ Provides a simple Resource Controller for monitoring the
+ total CPU consumed by the tasks in a cgroup
+
config RELAY
bool "Kernel->user space relay support (formerly relayfs)"
help
diff -puN kernel/Makefile~task-cgroupsv11-example-cpu-accounting-subsystem kernel/Makefile
--- a/kernel/Makefile~task-cgroupsv11-example-cpu-accounting-subsystem
+++ a/kernel/Makefile
@@ -40,6 +40,7 @@ obj-$(CONFIG_KEXEC) += kexec.o
obj-$(CONFIG_COMPAT) += compat.o
obj-$(CONFIG_CGROUPS) += cgroup.o
obj-$(CONFIG_CPUSETS) += cpuset.o
+obj-$(CONFIG_CGROUP_CPUACCT) += cpu_acct.o
obj-$(CONFIG_IKCONFIG) += configs.o
obj-$(CONFIG_STOP_MACHINE) += stop_machine.o
obj-$(CONFIG_AUDIT) += audit.o auditfilter.o
diff -puN /dev/null kernel/cpu_acct.c
--- /dev/null
+++ a/kernel/cpu_acct.c
@@ -0,0 +1,186 @@
+/*
+ * kernel/cpu_acct.c - CPU accounting cgroup subsystem
+ *
+ * Copyright (C) Google Inc, 2006
+ *
+ * Developed by Paul Menage (menage@google.com) and Balbir Singh
+ * (balbir@in.ibm.com)
+ *
+ */
+
+/*
+ * Example cgroup subsystem for reporting total CPU usage of tasks in a
+ * cgroup, along with percentage load over a time interval
+ */
+
+#include <linux/module.h>
+#include <linux/cgroup.h>
+#include <linux/fs.h>
+#include <linux/rcupdate.h>
+
+#include <asm/div64.h>
+
+struct cpuacct {
+ struct cgroup_subsys_state css;
+ spinlock_t lock;
+ /* total time used by this class */
+ cputime64_t time;
+
+ /* time when next load calculation occurs */
+ u64 next_interval_check;
+
+ /* time used in current period */
+ cputime64_t current_interval_time;
+
+ /* time used in last period */
+ cputime64_t last_interval_time;
+};
+
+struct cgroup_subsys cpuacct_subsys;
+
+static inline struct cpuacct *cgroup_ca(struct cgroup *cont)
+{
+ return container_of(cgroup_subsys_state(cont, cpuacct_subsys_id),
+ struct cpuacct, css);
+}
+
+static inline struct cpuacct *task_ca(struct task_struct *task)
+{
+ return container_of(task_subsys_state(task, cpuacct_subsys_id),
+ struct cpuacct, css);
+}
+
+#define INTERVAL (HZ * 10)
+
+static inline u64 next_interval_boundary(u64 now)
+{
+ /* calculate the next interval boundary beyond the
+ * current time */
+ do_div(now, INTERVAL);
+ return (now + 1) * INTERVAL;
+}
+
+static struct cgroup_subsys_state *cpuacct_create(
+ struct cgroup_subsys *ss, struct cgroup *cont)
+{
+ struct cpuacct *ca = kzalloc(sizeof(*ca), GFP_KERNEL);
+
+ if (!ca)
+ return ERR_PTR(-ENOMEM);
+ spin_lock_init(&ca->lock);
+ ca->next_interval_check = next_interval_boundary(get_jiffies_64());
+ return &ca->css;
+}
+
+static void cpuacct_destroy(struct cgroup_subsys *ss,
+ struct cgroup *cont)
+{
+ kfree(cgroup_ca(cont));
+}
+
+/* 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)) {
+ /* If it's been more than an interval since the last
+ * check, then catch up - the last interval must have
+ * been zero load */
+ ca->last_interval_time = 0;
+ ca->next_interval_check = next_interval_boundary(now);
+ } else {
+ /* If a steal takes the last interval time negative,
+ * then we just ignore it */
+ if ((s64)ca->current_interval_time > 0)
+ ca->last_interval_time = ca->current_interval_time;
+ else
+ ca->last_interval_time = 0;
+ ca->next_interval_check += INTERVAL;
+ }
+ ca->current_interval_time = 0;
+}
+
+static u64 cpuusage_read(struct cgroup *cont, struct cftype *cft)
+{
+ struct cpuacct *ca = cgroup_ca(cont);
+ u64 time;
+
+ spin_lock_irq(&ca->lock);
+ cpuusage_update(ca);
+ time = cputime64_to_jiffies64(ca->time);
+ spin_unlock_irq(&ca->lock);
+
+ /* Convert 64-bit jiffies to seconds */
+ time *= 1000;
+ do_div(time, HZ);
+ return time;
+}
+
+static u64 load_read(struct cgroup *cont, struct cftype *cft)
+{
+ struct cpuacct *ca = cgroup_ca(cont);
+ u64 time;
+
+ /* Find the time used in the previous interval */
+ spin_lock_irq(&ca->lock);
+ cpuusage_update(ca);
+ time = cputime64_to_jiffies64(ca->last_interval_time);
+ spin_unlock_irq(&ca->lock);
+
+ /* Convert time to a percentage, to give the load in the
+ * previous period */
+ time *= 100;
+ do_div(time, INTERVAL);
+
+ return time;
+}
+
+static struct cftype files[] = {
+ {
+ .name = "usage",
+ .read_uint = cpuusage_read,
+ },
+ {
+ .name = "load",
+ .read_uint = load_read,
+ }
+};
+
+static int cpuacct_populate(struct cgroup_subsys *ss, struct cgroup *cont)
+{
+ return cgroup_add_files(cont, ss, files, ARRAY_SIZE(files));
+}
+
+void cpuacct_charge(struct task_struct *task, cputime_t cputime)
+{
+
+ struct cpuacct *ca;
+ unsigned long flags;
+
+ if (!cpuacct_subsys.active)
+ return;
+ rcu_read_lock();
+ ca = task_ca(task);
+ if (ca) {
+ spin_lock_irqsave(&ca->lock, flags);
+ cpuusage_update(ca);
+ ca->time = cputime64_add(ca->time, cputime);
+ ca->current_interval_time =
+ cputime64_add(ca->current_interval_time, cputime);
+ spin_unlock_irqrestore(&ca->lock, flags);
+ }
+ rcu_read_unlock();
+}
+
+struct cgroup_subsys cpuacct_subsys = {
+ .name = "cpuacct",
+ .create = cpuacct_create,
+ .destroy = cpuacct_destroy,
+ .populate = cpuacct_populate,
+ .subsys_id = cpuacct_subsys_id,
+};
diff -puN kernel/sched.c~task-cgroupsv11-example-cpu-accounting-subsystem kernel/sched.c
--- a/kernel/sched.c~task-cgroupsv11-example-cpu-accounting-subsystem
+++ a/kernel/sched.c
@@ -51,6 +51,7 @@
#include <linux/cpu.h>
#include <linux/cpuset.h>
#include <linux/percpu.h>
+#include <linux/cpu_acct.h>
#include <linux/kthread.h>
#include <linux/seq_file.h>
#include <linux/sysctl.h>
@@ -3268,9 +3269,13 @@ 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)
@@ -3300,9 +3305,10 @@ 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);
@@ -3327,8 +3333,10 @@ 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);
+ }
}
/*
_
--
_______________________________________________
Containers mailing list
Containers@lists.linux-foundation.org
https://lists.linux-foundation.org/mailman/listinfo/containers
|
|
|
|
| [PATCH 07/29] task containersv11 shared container subsystem group arrays [message #20091 is a reply to message #20064] |
Tue, 11 September 2007 19:52 |
Paul Menage
Messages: 642
Registered: September 2006
|
Senior Member |
|
|
From: Paul Menage <menage@google.com>
Replace the struct css_set embedded in task_struct with a pointer; all tasks
that have the same set of memberships across all hierarchies will share a
css_set object, and will be linked via their css_sets field to the "tasks"
list_head in the css_set.
Assuming that many tasks share the same cgroup assignments, this reduces
overall space usage and keeps the size of the task_struct down (three pointers
added to task_struct compared to a non-cgroups kernel, no matter how many
subsystems are registered).
Signed-off-by: Paul Menage <menage@google.com>
Cc: Serge E. Hallyn <serue@us.ibm.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Dave Hansen <haveblue@us.ibm.com>
Cc: Balbir Singh <balbir@in.ibm.com>
Cc: Paul Jackson <pj@sgi.com>
Cc: Kirill Korotaev <dev@openvz.org>
Cc: Herbert Poetzl <herbert@13thfloor.at>
Cc: Srivatsa Vaddagiri <vatsa@in.ibm.com>
Cc: Cedric Le Goater <clg@fr.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---
Documentation/cgroups.txt | 14
include/linux/cgroup.h | 89 ++++
include/linux/sched.h | 33 -
kernel/cgroup.c | 606 ++++++++++++++++++++++++++++-----
kernel/fork.c | 1
5 files changed, 620 insertions(+), 123 deletions(-)
diff -puN Documentation/cgroups.txt~task-cgroupsv11-shared-cgroup-subsystem-group-arrays Documentation/cgroups.txt
--- a/Documentation/cgroups.txt~task-cgroupsv11-shared-cgroup-subsystem-group-arrays
+++ a/Documentation/cgroups.txt
@@ -176,7 +176,9 @@ Control Groups extends the kernel as follows
subsystem state is something that's expected to happen frequently
and in performance-critical code, whereas operations that require a
task's actual cgroup assignments (in particular, moving between
- cgroups) are less common.
+ cgroups) are less common. A linked list runs through the cg_list
+ field of each task_struct using the css_set, anchored at
+ css_set->tasks.
- A cgroup hierarchy filesystem can be mounted for browsing and
manipulation from user space.
@@ -252,6 +254,16 @@ linear search to locate an appropriate e
very efficient. A future version will use a hash table for better
performance.
+To allow access from a cgroup to the css_sets (and hence tasks)
+that comprise it, a set of cg_cgroup_link objects form a lattice;
+each cg_cgroup_link is linked into a list of cg_cgroup_links for
+a single cgroup on its cont_link_list field, and a list of
+cg_cgroup_links for a single css_set on its cg_link_list.
+
+Thus the set of tasks in a cgroup can be listed by iterating over
+each css_set that references the cgroup, and sub-iterating over
+each css_set's task set.
+
The use of a Linux virtual file system (vfs) to represent the
cgroup hierarchy provides for a familiar permission and name space
for cgroups, with a minimum of additional kernel code.
diff -puN include/linux/cgroup.h~task-cgroupsv11-shared-cgroup-subsystem-group-arrays include/linux/cgroup.h
--- a/include/linux/cgroup.h~task-cgroupsv11-shared-cgroup-subsystem-group-arrays
+++ a/include/linux/cgroup.h
@@ -27,10 +27,19 @@ extern void cgroup_lock(void);
extern void cgroup_unlock(void);
extern void cgroup_fork(struct task_struct *p);
extern void cgroup_fork_callbacks(struct task_struct *p);
+extern void cgroup_post_fork(struct task_struct *p);
extern void cgroup_exit(struct task_struct *p, int run_callbacks);
extern struct file_operations proc_cgroup_operations;
+/* Define the enumeration of all cgroup subsystems */
+#define SUBSYS(_x) _x ## _subsys_id,
+enum cgroup_subsys_id {
+#include <linux/cgroup_subsys.h>
+ CGROUP_SUBSYS_COUNT
+};
+#undef SUBSYS
+
/* Per-subsystem/per-cgroup state maintained by the system. */
struct cgroup_subsys_state {
/* The cgroup that this subsystem is attached to. Useful
@@ -97,6 +106,52 @@ struct cgroup {
struct cgroupfs_root *root;
struct cgroup *top_cgroup;
+
+ /*
+ * List of cg_cgroup_links pointing at css_sets with
+ * tasks in this cgroup. Protected by css_set_lock
+ */
+ struct list_head css_sets;
+};
+
+/* A css_set is a structure holding pointers to a set of
+ * cgroup_subsys_state objects. This saves space in the task struct
+ * object and speeds up fork()/exit(), since a single inc/dec and a
+ * list_add()/del() can bump the reference count on the entire
+ * cgroup set for a task.
+ */
+
+struct css_set {
+
+ /* Reference count */
+ struct kref ref;
+
+ /*
+ * List running through all cgroup groups. Protected by
+ * css_set_lock
+ */
+ struct list_head list;
+
+ /*
+ * List running through all tasks using this cgroup
+ * group. Protected by css_set_lock
+ */
+ struct list_head tasks;
+
+ /*
+ * List of cg_cgroup_link objects on link chains from
+ * cgroups referenced from this css_set. Protected by
+ * css_set_lock
+ */
+ struct list_head cg_links;
+
+ /*
+ * Set of subsystem states, one for each subsystem. This array
+ * is immutable after creation apart from the init_css_set
+ * during subsystem registration (at boot time).
+ */
+ struct cgroup_subsys_state *subsys[CGROUP_SUBSYS_COUNT];
+
};
/* struct cftype:
@@ -149,15 +204,7 @@ int cgroup_is_removed(const struct co
int cgroup_path(const struct cgroup *cont, char *buf, int buflen);
-int __cgroup_task_count(const struct cgroup *cont);
-static inline int cgroup_task_count(const struct cgroup *cont)
-{
- int task_count;
- rcu_read_lock();
- task_count = __cgroup_task_count(cont);
- rcu_read_unlock();
- return task_count;
-}
+int cgroup_task_count(const struct cgroup *cont);
/* Return true if the cgroup is a descendant of the current cgroup */
int cgroup_is_descendant(const struct cgroup *cont);
@@ -205,7 +252,7 @@ static inline struct cgroup_subsys_st
static inline struct cgroup_subsys_state *task_subsys_state(
struct task_struct *task, int subsys_id)
{
- return rcu_dereference(task->cgroups.subsys[subsys_id]);
+ return rcu_dereference(task->cgroups->subsys[subsys_id]);
}
static inline struct cgroup* task_cgroup(struct task_struct *task,
@@ -218,6 +265,27 @@ int cgroup_path(const struct containe
int cgroup_clone(struct task_struct *tsk, struct cgroup_subsys *ss);
+/* A cgroup_iter should be treated as an opaque object */
+struct cgroup_iter {
+ struct list_head *cg_link;
+ struct list_head *task;
+};
+
+/* To iterate across the tasks in a cgroup:
+ *
+ * 1) call cgroup_iter_start to intialize an iterator
+ *
+ * 2) call cgroup_iter_next() to retrieve member tasks until it
+ * returns NULL or until you want to end the iteration
+ *
+ * 3) call cgroup_iter_end() to destroy the iterator.
+ */
+void cgroup_iter_start(struct cgroup *cont, struct cgroup_iter *it);
+struct task_struct *cgroup_iter_next(struct cgroup *cont,
+ struct cgroup_iter *it);
+void cgroup_iter_end(struct cgroup *cont, struct cgroup_iter *it);
+
+
#else /* !CONFIG_CGROUPS */
static inline int cgroup_init_early(void) { return 0; }
@@ -225,6 +293,7 @@ static inline int cgroup_init(void) {
static inline void cgroup_init_smp(void) {}
static inline void cgroup_fork(struct task_struct *p) {}
static inline void cgroup_fork_callbacks(struct task_struct *p) {}
+static inline void cgroup_post_fork(struct task_struct *p) {}
static inline void cgroup_exit(struct task_struct *p, int callbacks) {}
static inline void cgroup_lock(void) {}
diff -puN include/linux/sched.h~task-cgroupsv11-shared-cgroup-subsystem-group-arrays include/linux/sched.h
--- a/include/linux/sched.h~task-cgroupsv11-shared-cgroup-subsystem-group-arrays
+++ a/include/linux/sched.h
@@ -861,34 +861,6 @@ struct sched_entity {
#endif
};
-#ifdef CONFIG_CGROUPS
-
-#define SUBSYS(_x) _x ## _subsys_id,
-enum cgroup_subsys_id {
-#include <linux/cgroup_subsys.h>
- CGROUP_SUBSYS_COUNT
-};
-#undef SUBSYS
-
-/* A css_set is a structure holding pointers to a set of
- * cgroup_subsys_state objects.
- */
-
-struct css_set {
-
- /* 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 cgroup, but result
- * in increased space usage if tasks are in wildly different
- * groupings across different hierarchies. This array is
- * immutable after creation */
- struct cgroup_subsys_state *subsys[CGROUP_SUBSYS_COUNT];
-
-};
-
-#endif /* CONFIG_CGROUPS */
-
struct task_struct {
volatile long state; /* -1 unrunnable, 0 runnable, >0 stopped */
void *stack;
@@ -1125,7 +1097,10 @@ struct task_struct {
int cpuset_mem_spread_rotor;
#endif
#ifdef CONFIG_CGROUPS
- struct css_set cgroups;
+ /* Control Group info protected by css_set_lock */
+ struct css_set *cgroups;
+ /* cg_list protected by css_set_lock and tsk->alloc_lock */
+ struct list_head cg_list;
#endif
#ifdef CONFIG_FUTEX
struct robust_list_head __user *robust_list;
diff -puN kernel/cgroup.c~task-cgroupsv11-shared-cgroup-subsystem-group-arrays kernel/cgroup.c
--- a/kernel/cgroup.c~task-cgroupsv11-shared-cgroup-subsystem-group-arrays
+++ a/kernel/cgroup.c
@@ -95,6 +95,7 @@ static struct cgroupfs_root rootnode;
/* The list of hierarchy roots */
static LIST_HEAD(roots);
+static int root_count;
/* dummytop is a shorthand for the dummy hierarchy's top cgroup */
#define dummytop (&rootnode.top_cgroup)
@@ -133,12 +134,49 @@ list_for_each_entry(_ss, &_root->subsys_
#define for_each_root(_root) \
list_for_each_entry(_root, &roots, root_list)
-/* Each task_struct has an embedded css_set, so the get/put
- * operation simply takes a reference count on all the cgroups
- * referenced by subsystems in this css_set. This can end up
- * multiple-counting some cgroups, but that's OK - the ref-count is
- * just a busy/not-busy indicator; ensuring that we only count each
- * cgroup once would require taking a global lock to ensure that no
+/* L
...
|
|
|
|
| [PATCH 01/29] task containersv11 basic task container framework [message #20092 is a reply to message #20064] |
Tue, 11 September 2007 19:52 |
Paul Menage
Messages: 642
Registered: September 2006
|
Senior Member |
|
|
From: Paul Menage <menage@google.com>
Generic Process Control Groups
--------------------------
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
cgroups, 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.
This patchset provides a framework for tracking and grouping processes
into arbitrary "cgroups" and assigning arbitrary state to those
groupings, in order to control the behaviour of the cgroup as an
aggregate.
The intention is that the various resource management and
virtualization/cgroup efforts can also become task cgroup
clients, with the result that:
- the userspace APIs are (somewhat) normalised
- it's easier to test 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
This patch:
Add the main task cgroups framework - the cgroup filesystem, and the
basic structures for tracking membership and associating subsystem state
objects to tasks.
Signed-off-by: Paul Menage <menage@google.com>
Cc: Serge E. Hallyn <serue@us.ibm.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Dave Hansen <haveblue@us.ibm.com>
Cc: Balbir Singh <balbir@in.ibm.com>
Cc: Paul Jackson <pj@sgi.com>
Cc: Kirill Korotaev <dev@openvz.org>
Cc: Herbert Poetzl <herbert@13thfloor.at>
Cc: Srivatsa Vaddagiri <vatsa@in.ibm.com>
Cc: Cedric Le Goater <clg@fr.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---
Documentation/cgroups.txt | 526 ++++++++++++
include/linux/cgroup.h | 214 +++++
include/linux/cgroup_subsys.h | 10
include/linux/magic.h | 1
include/linux/sched.h | 34
init/Kconfig | 8
init/main.c | 3
kernel/Makefile | 1
kernel/cgroup.c | 1199 +++++++++++++++++++++++++++++
9 files changed, 1995 insertions(+), 1 deletion(-)
diff -puN /dev/null Documentation/cgroups.txt
--- /dev/null
+++ a/Documentation/cgroups.txt
@@ -0,0 +1,526 @@
+ CGROUPS
+ -------
+
+Written by Paul Menage <menage@google.com> 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 <pj@sgi.com>
+Modified by Christoph Lameter <clameter@sgi.com>
+
+CONTENTS:
+=========
+
+1. Control Groups
+ 1.1 What are cgroups ?
+ 1.2 Why are cgroups needed ?
+ 1.3 How are cgroups implemented ?
+ 1.4 What does notify_on_release do ?
+ 1.5 How do I use cgroups ?
+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. Control Groups
+==========
+
+1.1 What are cgroups ?
+----------------------
+
+Control Groups provide a mechanism for aggregating/partitioning sets of
+tasks, and all their future children, into hierarchical groups with
+specialized behaviour.
+
+Definitions:
+
+A *cgroup* 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 cgroups to treat groups of tasks in
+particular ways. A subsystem is typically a "resource controller" that
+schedules a resource or applies per-cgroup 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 cgroups arranged in a tree, such that
+every task in the system is in exactly one of the cgroups in the
+hierarchy, and a set of subsystems; each subsystem has system-specific
+state attached to each cgroup in the hierarchy. Each hierarchy has
+an instance of the cgroup virtual filesystem associated with it.
+
+At any one time there may be multiple active hierachies of task
+cgroups. Each hierarchy is a partition of all tasks in the system.
+
+User level code may create and destroy cgroups by name in an
+instance of the cgroup virtual file system, specify and query to
+which cgroup a task is assigned, and list the task pids assigned to
+a cgroup. Those creations and assignments only affect the hierarchy
+associated with that instance of the cgroup file system.
+
+On their own, the only use for cgroups is for simple job
+tracking. The intention is that other subsystems hook into the generic
+cgroup support to provide new attributes for cgroups, such as
+accounting/limiting the resources which processes in a cgroup 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 cgroup.
+
+1.2 Why are cgroups 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 (cgroup) as their parent process.
+
+The kernel cgroup 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 cgroups 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
+cgroups.
+
+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 vatsa@in.ibm.com)
+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 cgroup for every browser launched and associate it with
+approp network and other resource class. This may lead to
+proliferation of such cgroups.
+
+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 cgroup into
+multiple separate ones and then associate the new cgroups with the
+new resource classes.
+
+
+
+1.3 How are cgroups implemented ?
+---------------------------------
+
+Control Groups extends the kernel as follows:
+
+ - Each task in the system has a reference-counted pointer to a
+ css_set.
+
+ - A css_set contains a set of reference-counted pointers to
+ cgroup_subsys_state objects, one for each cgroup subsystem
+ registered in the system. There is no direct link from a task to
+ the cgroup of which it's a member in each hierarchy, but this
+ can be determined by following pointers through the
+ cgroup_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 cgroup assignments (in particular, moving between
+ cgroups) are less common.
+
+ - A cgroup hierarchy filesystem can be mounted for browsing and
+ manipulation from user space.
+
+ - You
...
|
|
|
|
| [PATCH 11/29] task containersv11 make cpusets a client of containers [message #20093 is a reply to message #20064] |
Tue, 11 September 2007 19:52 |
Paul Menage
Messages: 642
Registered: September 2006
|
Senior Member |
|
|
From: Paul Menage <menage@google.com>
Remove the filesystem support logic from the cpusets system and makes cpusets
a cgroup subsystem
The "cpuset" filesystem becomes a dummy filesystem; attempts to mount it get
passed through to the cgroup filesystem with the appropriate options to
emulate the old cpuset filesystem behaviour.
Signed-off-by: Paul Menage <menage@google.com>
Cc: Serge E. Hallyn <serue@us.ibm.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Dave Hansen <haveblue@us.ibm.com>
Cc: Balbir Singh <balbir@in.ibm.com>
Cc: Paul Jackson <pj@sgi.com>
Cc: Kirill Korotaev <dev@openvz.org>
Cc: Herbert Poetzl <herbert@13thfloor.at>
Cc: Srivatsa Vaddagiri <vatsa@in.ibm.com>
Cc: Cedric Le Goater <clg@fr.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---
Documentation/cpusets.txt | 91 +-
fs/proc/base.c | 4
include/linux/cgroup_subsys.h | 6
include/linux/cpuset.h | 12
include/linux/mempolicy.h | 12
include/linux/sched.h | 3
init/Kconfig | 7
kernel/cpuset.c | 1192 +++++------------------------
kernel/exit.c | 2
kernel/fork.c | 3
mm/mempolicy.c | 2
11 files changed, 278 insertions(+), 1056 deletions(-)
diff -puN Documentation/cpusets.txt~task-cgroupsv11-make-cpusets-a-client-of-cgroups Documentation/cpusets.txt
--- a/Documentation/cpusets.txt~task-cgroupsv11-make-cpusets-a-client-of-cgroups
+++ a/Documentation/cpusets.txt
@@ -7,6 +7,7 @@ Written by Simon.Derr@bull.net
Portions Copyright (c) 2004-2006 Silicon Graphics, Inc.
Modified by Paul Jackson <pj@sgi.com>
Modified by Christoph Lameter <clameter@sgi.com>
+Modified by Paul Menage <menage@google.com>
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
@@ -44,18 +44,19 @@ 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 cgroup subsystem described in
+Documentation/cgroup.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 cgroup
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
@@ -115,7 +116,7 @@ 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 cgroup 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
@@ -145,15 +146,10 @@ 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 "cgroup" 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)
@@ -163,16 +159,15 @@ 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 cgroup file system
+containing (on top of the standard cgroup 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:
@@ -237,21 +232,7 @@ 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
@@ -308,7 +289,7 @@ 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
@@ -379,7 +360,7 @@ 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
@@ -469,7 +450,7 @@ 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 cgroup -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.
@@ -481,7 +462,7 @@ 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 cgroup -ocpuset cpuset /dev/cpuset
cd /dev/cpuset
mkdir Charlie
cd Charlie
@@ -513,7 +494,7 @@ Creating, modifying, using the cpusets c
virtual filesystem.
To mount it, type:
-# mount -t cpuset none /dev/cpuset
+# mount -t cgroup -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
@@ -556,6 +537,18 @@ 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 cgroup filesystem.
+
+The command
+
+mount -t cpuset X /dev/cpuset
+
+is equivalent to
+
+mount -t cgroup -ocpuset X /dev/cpuset
+echo "/sbin/cpuset_release_agent" > /dev/cpuset/release_agent
+
2.2 Adding/removing cpus
------------------------
diff -puN fs/proc/base.c~task-cgroupsv11-make-cpusets-a-client-of-cgroups fs/proc/base.c
--- a/fs/proc/base.c~task-cgroupsv11-make-cpusets-a-client-of-cgroups
+++ a/fs/proc/base.c
@@ -
...
|
|
|
|
| [PATCH 21/29] memory controller containers setup v7 [message #20094 is a reply to message #20064] |
Tue, 11 September 2007 19:53 |
Paul Menage
Messages: 642
Registered: September 2006
|
Senior Member |
|
|
From: Balbir Singh <balbir@linux.vnet.ibm.com>
Setup the memory cgroup and add basic hooks and controls to integrate
and work with the cgroup.
Signed-off-by: Balbir Singh <balbir@linux.vnet.ibm.com>
Cc: Pavel Emelianov <xemul@openvz.org>
Cc: Paul Menage <menage@google.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Cc: Kirill Korotaev <dev@sw.ru>
Cc: Herbert Poetzl <herbert@13thfloor.at>
Cc: David Rientjes <rientjes@google.com>
Cc: Vaidyanathan Srinivasan <svaidy@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---
include/linux/cgroup_subsys.h | 6 +
include/linux/memcontrol.h | 21 ++++
init/Kconfig | 7 +
mm/Makefile | 1
mm/memcontrol.c | 127 +++++++++++++++++++++++++++++
5 files changed, 162 insertions(+)
diff -puN include/linux/cgroup_subsys.h~memory-controller-cgroups-setup-v7 include/linux/cgroup_subsys.h
--- a/include/linux/cgroup_subsys.h~memory-controller-cgroups-setup-v7
+++ a/include/linux/cgroup_subsys.h
@@ -30,3 +30,9 @@ SUBSYS(ns)
#endif
/* */
+
+#ifdef CONFIG_CGROUP_MEM_CONT
+SUBSYS(mem_cgroup)
+#endif
+
+/* */
diff -puN /dev/null include/linux/memcontrol.h
--- /dev/null
+++ a/include/linux/memcontrol.h
@@ -0,0 +1,21 @@
+/* memcontrol.h - Memory Controller
+ *
+ * Copyright IBM Corporation, 2007
+ * Author Balbir Singh <balbir@linux.vnet.ibm.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ */
+
+#ifndef _LINUX_MEMCONTROL_H
+#define _LINUX_MEMCONTROL_H
+
+#endif /* _LINUX_MEMCONTROL_H */
+
diff -puN init/Kconfig~memory-controller-cgroups-setup-v7 init/Kconfig
--- a/init/Kconfig~memory-controller-cgroups-setup-v7
+++ a/init/Kconfig
@@ -346,6 +346,13 @@ 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 CGROUP_MEM_CONT
+ bool "Memory controller for cgroups"
+ depends on CGROUPS && RESOURCE_COUNTERS
+ help
+ Provides a memory controller that manages both page cache and
+ RSS memory.
+
config PROC_PID_CPUSET
bool "Include legacy /proc/<pid>/cpuset file"
depends on CPUSETS
diff -puN mm/Makefile~memory-controller-cgroups-setup-v7 mm/Makefile
--- a/mm/Makefile~memory-controller-cgroups-setup-v7
+++ a/mm/Makefile
@@ -30,4 +30,5 @@ obj-$(CONFIG_FS_XIP) += filemap_xip.o
obj-$(CONFIG_MIGRATION) += migrate.o
obj-$(CONFIG_SMP) += allocpercpu.o
obj-$(CONFIG_QUICKLIST) += quicklist.o
+obj-$(CONFIG_CGROUP_MEM_CONT) += memcontrol.o
diff -puN /dev/null mm/memcontrol.c
--- /dev/null
+++ a/mm/memcontrol.c
@@ -0,0 +1,127 @@
+/* memcontrol.c - Memory Controller
+ *
+ * Copyright IBM Corporation, 2007
+ * Author Balbir Singh <balbir@linux.vnet.ibm.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/res_counter.h>
+#include <linux/memcontrol.h>
+#include <linux/cgroup.h>
+
+struct cgroup_subsys mem_cgroup_subsys;
+
+/*
+ * The memory controller data structure. The memory controller controls both
+ * page cache and RSS per cgroup. We would eventually like to provide
+ * statistics based on the statistics developed by Rik Van Riel for clock-pro,
+ * to help the administrator determine what knobs to tune.
+ *
+ * TODO: Add a water mark for the memory controller. Reclaim will begin when
+ * we hit the water mark.
+ */
+struct mem_cgroup {
+ struct cgroup_subsys_state css;
+ /*
+ * the counter to account for memory usage
+ */
+ struct res_counter res;
+};
+
+/*
+ * A page_cgroup page is associated with every page descriptor. The
+ * page_cgroup helps us identify information about the cgroup
+ */
+struct page_cgroup {
+ struct list_head lru; /* per cgroup LRU list */
+ struct page *page;
+ struct mem_cgroup *mem_cgroup;
+};
+
+
+static inline
+struct mem_cgroup *mem_cgroup_from_cont(struct cgroup *cont)
+{
+ return container_of(cgroup_subsys_state(cont,
+ mem_cgroup_subsys_id), struct mem_cgroup,
+ css);
+}
+
+static ssize_t mem_cgroup_read(struct cgroup *cont, struct cftype *cft,
+ struct file *file, char __user *userbuf, size_t nbytes,
+ loff_t *ppos)
+{
+ return res_counter_read(&mem_cgroup_from_cont(cont)->res,
+ cft->private, userbuf, nbytes, ppos);
+}
+
+static ssize_t mem_cgroup_write(struct cgroup *cont, struct cftype *cft,
+ struct file *file, const char __user *userbuf,
+ size_t nbytes, loff_t *ppos)
+{
+ return res_counter_write(&mem_cgroup_from_cont(cont)->res,
+ cft->private, userbuf, nbytes, ppos);
+}
+
+static struct cftype mem_cgroup_files[] = {
+ {
+ .name = "usage",
+ .private = RES_USAGE,
+ .read = mem_cgroup_read,
+ },
+ {
+ .name = "limit",
+ .private = RES_LIMIT,
+ .write = mem_cgroup_write,
+ .read = mem_cgroup_read,
+ },
+ {
+ .name = "failcnt",
+ .private = RES_FAILCNT,
+ .read = mem_cgroup_read,
+ },
+};
+
+static struct cgroup_subsys_state *
+mem_cgroup_create(struct cgroup_subsys *ss, struct cgroup *cont)
+{
+ struct mem_cgroup *mem;
+
+ mem = kzalloc(sizeof(struct mem_cgroup), GFP_KERNEL);
+ if (!mem)
+ return -ENOMEM;
+
+ res_counter_init(&mem->res);
+ return &mem->css;
+}
+
+static void mem_cgroup_destroy(struct cgroup_subsys *ss,
+ struct cgroup *cont)
+{
+ kfree(mem_cgroup_from_cont(cont));
+}
+
+static int mem_cgroup_populate(struct cgroup_subsys *ss,
+ struct cgroup *cont)
+{
+ return cgroup_add_files(cont, ss, mem_cgroup_files,
+ ARRAY_SIZE(mem_cgroup_files));
+}
+
+struct cgroup_subsys mem_cgroup_subsys = {
+ .name = "memory",
+ .subsys_id = mem_cgroup_subsys_id,
+ .create = mem_cgroup_create,
+ .destroy = mem_cgroup_destroy,
+ .populate = mem_cgroup_populate,
+ .early_init = 0,
+};
_
--
_______________________________________________
Containers mailing list
Containers@lists.linux-foundation.org
https://lists.linux-foundation.org/mailman/listinfo/containers
|
|
|
|
| [PATCH 19/29] memory controller resource counters v7 [message #20095 is a reply to message #20064] |
Tue, 11 September 2007 19:52 |
Paul Menage
Messages: 642
Registered: September 2006
|
Senior Member |
|
|
From: Pavel Emelianov <xemul@openvz.org>
Introduce generic structures and routines for resource accounting.
Each resource accounting cgroup is supposed to aggregate it,
cgroup_subsystem_state and its resource-specific members within.
Signed-off-by: Pavel Emelianov <xemul@openvz.org>
Signed-off-by: Balbir Singh <balbir@linux.vnet.ibm.com>
Cc: Paul Menage <menage@google.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Cc: Kirill Korotaev <dev@sw.ru>
Cc: Herbert Poetzl <herbert@13thfloor.at>
Cc: David Rientjes <rientjes@google.com>
Cc: Vaidyanathan Srinivasan <svaidy@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---
include/linux/res_counter.h | 102 ++++++++++++++++++++++++++++
init/Kconfig | 7 +
kernel/Makefile | 1
kernel/res_counter.c | 120 ++++++++++++++++++++++++++++++++++
4 files changed, 230 insertions(+)
diff -puN /dev/null include/linux/res_counter.h
--- /dev/null
+++ a/include/linux/res_counter.h
@@ -0,0 +1,102 @@
+#ifndef __RES_COUNTER_H__
+#define __RES_COUNTER_H__
+
+/*
+ * Resource Counters
+ * Contain common data types and routines for resource accounting
+ *
+ * Copyright 2007 OpenVZ SWsoft Inc
+ *
+ * Author: Pavel Emelianov <xemul@openvz.org>
+ *
+ */
+
+#include <linux/cgroup.h>
+
+/*
+ * The core object. the cgroup that wishes to account for some
+ * resource may include this counter into its structures and use
+ * the helpers described beyond
+ */
+
+struct res_counter {
+ /*
+ * the current resource consumption level
+ */
+ unsigned long usage;
+ /*
+ * the limit that usage cannot exceed
+ */
+ unsigned long limit;
+ /*
+ * the number of unsuccessful attempts to consume the resource
+ */
+ unsigned long failcnt;
+ /*
+ * the lock to protect all of the above.
+ * the routines below consider this to be IRQ-safe
+ */
+ spinlock_t lock;
+};
+
+/*
+ * Helpers to interact with userspace
+ * res_counter_read/_write - put/get the specified fields from the
+ * res_counter struct to/from the user
+ *
+ * @counter: the counter in question
+ * @member: the field to work with (see RES_xxx below)
+ * @buf: the buffer to opeate on,...
+ * @nbytes: its size...
+ * @pos: and the offset.
+ */
+
+ssize_t res_counter_read(struct res_counter *counter, int member,
+ const char __user *buf, size_t nbytes, loff_t *pos);
+ssize_t res_counter_write(struct res_counter *counter, int member,
+ const char __user *buf, size_t nbytes, loff_t *pos);
+
+/*
+ * the field descriptors. one for each member of res_counter
+ */
+
+enum {
+ RES_USAGE,
+ RES_LIMIT,
+ RES_FAILCNT,
+};
+
+/*
+ * helpers for accounting
+ */
+
+void res_counter_init(struct res_counter *counter);
+
+/*
+ * charge - try to consume more resource.
+ *
+ * @counter: the counter
+ * @val: the amount of the resource. each controller defines its own
+ * units, e.g. numbers, bytes, Kbytes, etc
+ *
+ * returns 0 on success and <0 if the counter->usage will exceed the
+ * counter->limit _locked call expects the counter->lock to be taken
+ */
+
+int res_counter_charge_locked(struct res_counter *counter, unsigned long val);
+int res_counter_charge(struct res_counter *counter, unsigned long val);
+
+/*
+ * uncharge - tell that some portion of the resource is released
+ *
+ * @counter: the counter
+ * @val: the amount of the resource
+ *
+ * these calls check for usage underflow and show a warning on the console
+ * _locked call expects the counter->lock to be taken
+ */
+
+void res_counter_uncharge_locked(struct res_counter *counter, unsigned long val);
+void res_counter_uncharge(struct res_counter *counter, unsigned long val);
+
+#endif
diff -puN init/Kconfig~memory-controller-resource-counters-v7 init/Kconfig
--- a/init/Kconfig~memory-controller-resource-counters-v7
+++ a/init/Kconfig
@@ -319,6 +319,13 @@ config CPUSETS
Say N if unsure.
+config RESOURCE_COUNTERS
+ bool "Resource counters"
+ help
+ This option enables controller independent resource accounting
+ infrastructure that works with cgroups
+ depends on CGROUPS
+
config SYSFS_DEPRECATED
bool "Create deprecated sysfs files"
default y
diff -puN kernel/Makefile~memory-controller-resource-counters-v7 kernel/Makefile
--- a/kernel/Makefile~memory-controller-resource-counters-v7
+++ a/kernel/Makefile
@@ -59,6 +59,7 @@ obj-$(CONFIG_RELAY) += relay.o
obj-$(CONFIG_SYSCTL) += utsname_sysctl.o
obj-$(CONFIG_TASK_DELAY_ACCT) += delayacct.o
obj-$(CONFIG_TASKSTATS) += taskstats.o tsacct.o
+obj-$(CONFIG_RESOURCE_COUNTERS) += res_counter.o
ifneq ($(CONFIG_SCHED_NO_NO_OMIT_FRAME_POINTER),y)
# According to Alan Modra <alan@linuxcare.com.au>, the -fno-omit-frame-pointer is
diff -puN /dev/null kernel/res_counter.c
--- /dev/null
+++ a/kernel/res_counter.c
@@ -0,0 +1,120 @@
+/*
+ * resource cgroups
+ *
+ * Copyright 2007 OpenVZ SWsoft Inc
+ *
+ * Author: Pavel Emelianov <xemul@openvz.org>
+ *
+ */
+
+#include <linux/types.h>
+#include <linux/parser.h>
+#include <linux/fs.h>
+#include <linux/res_counter.h>
+#include <linux/uaccess.h>
+
+void res_counter_init(struct res_counter *counter)
+{
+ spin_lock_init(&counter->lock);
+ counter->limit = (unsigned long)LONG_MAX;
+}
+
+int res_counter_charge_locked(struct res_counter *counter, unsigned long val)
+{
+ if (counter->usage > (counter->limit - val)) {
+ counter->failcnt++;
+ return -ENOMEM;
+ }
+
+ counter->usage += val;
+ return 0;
+}
+
+int res_counter_charge(struct res_counter *counter, unsigned long val)
+{
+ int ret;
+ unsigned long flags;
+
+ spin_lock_irqsave(&counter->lock, flags);
+ ret = res_counter_charge_locked(counter, val);
+ spin_unlock_irqrestore(&counter->lock, flags);
+ return ret;
+}
+
+void res_counter_uncharge_locked(struct res_counter *counter, unsigned long val)
+{
+ if (WARN_ON(counter->usage < val))
+ val = counter->usage;
+
+ counter->usage -= val;
+}
+
+void res_counter_uncharge(struct res_counter *counter, unsigned long val)
+{
+ unsigned long flags;
+
+ spin_lock_irqsave(&counter->lock, flags);
+ res_counter_uncharge_locked(counter, val);
+ spin_unlock_irqrestore(&counter->lock, flags);
+}
+
+
+static inline unsigned long *res_counter_member(struct res_counter *counter,
+ int member)
+{
+ switch (member) {
+ case RES_USAGE:
+ return &counter->usage;
+ case RES_LIMIT:
+ return &counter->limit;
+ case RES_FAILCNT:
+ return &counter->failcnt;
+ };
+
+ BUG();
+ return NULL;
+}
+
+ssize_t res_counter_read(struct res_counter *counter, int member,
+ const char __user *userbuf, size_t nbytes, loff_t *pos)
+{
+ unsigned long *val;
+ char buf[64], *s;
+
+ s = buf;
+ val = res_counter_member(counter, member);
+ s += sprintf(s, "%lu\n", *val);
+ return simple_read_from_buffer((void __user *)userbuf, nbytes,
+ pos, buf, s - buf);
+}
+
+ssize_t res_counter_write(struct res_counter *counter, int member,
+ const char __user *userbuf, size_t nbytes, loff_t *pos)
+{
+ int ret;
+ char *buf, *end;
+ unsigned long tmp, *val;
+
+ buf = kmalloc(nbytes + 1, GFP_KERNEL);
+ ret = -ENOMEM;
+ if (buf == NULL)
+ goto out;
+
+ buf[nbytes] = '\0';
+ ret = -EFAULT;
+ if (copy_from_user(buf, userbuf, nbytes))
+ goto out_free;
+
+ ret = -EINVAL;
+ tmp = simple_strtoul(buf, &end, 10);
+ if (*end != '\0')
+ goto out_free;
+
+ val = res_counter_member(counter, member);
+ *val = tmp;
+ ret = nbytes;
+out_free:
+ kfree(buf);
+out:
+ return ret;
+}
_
--
_______________________________________________
Containers mailing list
Containers@lists.linux-foundation.org
https://lists.linux-foundation.org/mailman/listinfo/containers
|
|
|
|
| Re: [PATCH 23/29] memory controller memory accounting v7 [message #20174 is a reply to message #20078] |
Wed, 12 September 2007 20:56 |
Peter Zijlstra
Messages: 61
Registered: September 2006
|
Member |
|
|
> From: Balbir Singh <balbir@linux.vnet.ibm.com>
> void page_assign_page_cgroup(struct page *page, struct page_cgroup *pc)
> {
> - page->page_cgroup = (unsigned long)pc;
> + int locked;
> +
> + /*
> + * While resetting the page_cgroup we might not hold the
> + * page_cgroup lock. free_hot_cold_page() is an example
> + * of such a scenario
> + */
> + if (pc)
> + VM_BUG_ON(!page_cgroup_locked(page));
> + locked = (page->page_cgroup & PAGE_CGROUP_LOCK);
> + page->page_cgroup = ((unsigned long)pc | locked);
> }
This looks a bit odd, why not write:
locked = page_cgroup_locked(page);
if (pc)
VM_BUG_ON(!locked)
> +/*
> + * Charge the memory controller for page usage.
> + * Return
> + * 0 if the charge was successful
> + * < 0 if the cgroup is over its limit
> + */
> +int mem_cgroup_charge(struct page *page, struct mm_struct *mm)
> +{
> + struct mem_cgroup *mem;
> + struct page_cgroup *pc, *race_pc;
> +
> + /*
> + * Should page_cgroup's go to their own slab?
> + * One could optimize the performance of the charging routine
> + * by saving a bit in the page_flags and using it as a lock
> + * to see if the cgroup page already has a page_cgroup associated
> + * with it
> + */
> + lock_page_cgroup(page);
> + pc = page_get_page_cgroup(page);
> + /*
> + * The page_cgroup exists and the page has already been accounted
> + */
> + if (pc) {
> + atomic_inc(&pc->ref_cnt);
> + goto done;
> + }
> +
> + unlock_page_cgroup(page);
> +
> + pc = kzalloc(sizeof(struct page_cgroup), GFP_KERNEL);
> + if (pc == NULL)
> + goto err;
> +
> + rcu_read_lock();
> + /*
> + * We always charge the cgroup the mm_struct belongs to
> + * the mm_struct's mem_cgroup changes on task migration if the
> + * thread group leader migrates. It's possible that mm is not
> + * set, if so charge the init_mm (happens for pagecache usage).
> + */
> + if (!mm)
> + mm = &init_mm;
> +
> + mem = rcu_dereference(mm->mem_cgroup);
> + /*
> + * For every charge from the cgroup, increment reference
> + * count
> + */
> + css_get(&mem->css);
> + rcu_read_unlock();
> +
> + /*
> + * If we created the page_cgroup, we should free it on exceeding
> + * the cgroup limit.
> + */
> + if (res_counter_charge(&mem->res, 1)) {
> + css_put(&mem->css);
> + goto free_pc;
> + }
> +
> + lock_page_cgroup(page);
> + /*
> + * Check if somebody else beat us to allocating the page_cgroup
> + */
> + race_pc = page_get_page_cgroup(page);
> + if (race_pc) {
> + kfree(pc);
> + pc = race_pc;
> + atomic_inc(&pc->ref_cnt);
This inc
> + res_counter_uncharge(&mem->res, 1);
> + css_put(&mem->css);
> + goto done;
> + }
> +
> + atomic_set(&pc->ref_cnt, 1);
combined with this set make me wonder...
> + pc->mem_cgroup = mem;
> + pc->page = page;
> + page_assign_page_cgroup(page, pc);
> +
> +done:
> + unlock_page_cgroup(page);
> + return 0;
> +free_pc:
> + kfree(pc);
> + return -ENOMEM;
> +err:
> + unlock_page_cgroup(page);
> + return -ENOMEM;
> +}
> @@ -2161,6 +2184,9 @@ static int do_anonymous_page(struct mm_s
> if (!page)
> goto oom;
>
> + if (mem_cgroup_charge(page, mm))
> + goto oom_free_page;
> +
> entry = mk_pte(page, vma->vm_page_prot);
> entry = maybe_mkwrite(pte_mkdirty(entry), vma);
>
whitespace damage
_______________________________________________
Containers mailing list
Containers@lists.linux-foundation.org
https://lists.linux-foundation.org/mailman/listinfo/containers
|
|
|
|
| Re: [PATCH 23/29] memory controller memory accounting v7 [message #20206 is a reply to message #20174] |
Thu, 13 September 2007 09:49 |
Balbir Singh
Messages: 491
Registered: August 2006
|
Senior Member |
|
|
Peter Zijlstra wrote:
>> From: Balbir Singh <balbir@linux.vnet.ibm.com>
>
>> void page_assign_page_cgroup(struct page *page, struct page_cgroup *pc)
>> {
>> - page->page_cgroup = (unsigned long)pc;
>> + int locked;
>> +
>> + /*
>> + * While resetting the page_cgroup we might not hold the
>> + * page_cgroup lock. free_hot_cold_page() is an example
>> + * of such a scenario
>> + */
>> + if (pc)
>> + VM_BUG_ON(!page_cgroup_locked(page));
>> + locked = (page->page_cgroup & PAGE_CGROUP_LOCK);
>> + page->page_cgroup = ((unsigned long)pc | locked);
>> }
>
> This looks a bit odd, why not write:
>
> locked = page_cgroup_locked(page);
> if (pc)
> VM_BUG_ON(!locked)
>
Sure, we could write it this way or
VM_BUG_ON(pc && !locked)
>> +/*
>> + * Charge the memory controller for page usage.
>> + * Return
>> + * 0 if the charge was successful
>> + * < 0 if the cgroup is over its limit
>> + */
>> +int mem_cgroup_charge(struct page *page, struct mm_struct *mm)
>> +{
>> + struct mem_cgroup *mem;
>> + struct page_cgroup *pc, *race_pc;
>> +
>> + /*
>> + * Should page_cgroup's go to their own slab?
>> + * One could optimize the performance of the charging routine
>> + * by saving a bit in the page_flags and using it as a lock
>> + * to see if the cgroup page already has a page_cgroup associated
>> + * with it
>> + */
>> + lock_page_cgroup(page);
>> + pc = page_get_page_cgroup(page);
>> + /*
>> + * The page_cgroup exists and the page has already been accounted
>> + */
>> + if (pc) {
>> + atomic_inc(&pc->ref_cnt);
>> + goto done;
>> + }
>> +
>> + unlock_page_cgroup(page);
>> +
>> + pc = kzalloc(sizeof(struct page_cgroup), GFP_KERNEL);
>> + if (pc == NULL)
>> + goto err;
>> +
>> + rcu_read_lock();
>> + /*
>> + * We always charge the cgroup the mm_struct belongs to
>> + * the mm_struct's mem_cgroup changes on task migration if the
>> + * thread group leader migrates. It's possible that mm is not
>> + * set, if so charge the init_mm (happens for pagecache usage).
>> + */
>> + if (!mm)
>> + mm = &init_mm;
>> +
>> + mem = rcu_dereference(mm->mem_cgroup);
>> + /*
>> + * For every charge from the cgroup, increment reference
>> + * count
>> + */
>> + css_get(&mem->css);
>> + rcu_read_unlock();
>> +
>> + /*
>> + * If we created the page_cgroup, we should free it on exceeding
>> + * the cgroup limit.
>> + */
>> + if (res_counter_charge(&mem->res, 1)) {
>> + css_put(&mem->css);
>> + goto free_pc;
>> + }
>> +
>> + lock_page_cgroup(page);
>> + /*
>> + * Check if somebody else beat us to allocating the page_cgroup
>> + */
>> + race_pc = page_get_page_cgroup(page);
>> + if (race_pc) {
>> + kfree(pc);
>> + pc = race_pc;
>> + atomic_inc(&pc->ref_cnt);
>
> This inc
>
>> + res_counter_uncharge(&mem->res, 1);
>> + css_put(&mem->css);
>> + goto done;
>> + }
>> +
>> + atomic_set(&pc->ref_cnt, 1);
>
> combined with this set make me wonder...
>
I am not sure I understand this comment.
>> + pc->mem_cgroup = mem;
>> + pc->page = page;
>> + page_assign_page_cgroup(page, pc);
>> +
>> +done:
>> + unlock_page_cgroup(page);
>> + return 0;
>> +free_pc:
>> + kfree(pc);
>> + return -ENOMEM;
>> +err:
>> + unlock_page_cgroup(page);
>> + return -ENOMEM;
>> +}
>
>
>
>> @@ -2161,6 +2184,9 @@ static int do_anonymous_page(struct mm_s
>> if (!page)
>> goto oom;
>>
>> + if (mem_cgroup_charge(page, mm))
>> + goto oom_free_page;
>> +
>> entry = mk_pte(page, vma->vm_page_prot);
>> entry = maybe_mkwrite(pte_mkdirty(entry), vma);
>>
>
> whitespace damage
>
Yes, it's already been fixed in Andrew's tree. Paul could you
please pull in the those fixes as well? They are not in a -mm
tree, but you can find them on mm-commits.
--
Warm Regards,
Balbir Singh
Linux Technology Center
IBM, ISTL
_______________________________________________
Containers mailing list
Containers@lists.linux-foundation.org
https://lists.linux-foundation.org/mailman/listinfo/containers
|
|
|
|
| Re: [PATCH 23/29] memory controller memory accounting v7 [message #20208 is a reply to message #20206] |
Thu, 13 September 2007 10:18 |
Peter Zijlstra
Messages: 61
Registered: September 2006
|
Member |
|
|
On Thu, 2007-09-13 at 15:19 +0530, Balbir Singh wrote:
> VM_BUG_ON(pc && !locked)
Even better :-)
> >> +/*
> >> + * Charge the memory controller for page usage.
> >> + * Return
> >> + * 0 if the charge was successful
> >> + * < 0 if the cgroup is over its limit
> >> + */
> >> +int mem_cgroup_charge(struct page *page, struct mm_struct *mm)
> >> +{
> >> + struct mem_cgroup *mem;
> >> + struct page_cgroup *pc, *race_pc;
> >> +
> >> + /*
> >> + * Should page_cgroup's go to their own slab?
> >> + * One could optimize the performance of the charging routine
> >> + * by saving a bit in the page_flags and using it as a lock
> >> + * to see if the cgroup page already has a page_cgroup associated
> >> + * with it
> >> + */
> >> + lock_page_cgroup(page);
> >> + pc = page_get_page_cgroup(page);
> >> + /*
> >> + * The page_cgroup exists and the page has already been accounted
> >> + */
> >> + if (pc) {
> >> + atomic_inc(&pc->ref_cnt);
> >> + goto done;
> >> + }
> >> +
> >> + unlock_page_cgroup(page);
> >> +
> >> + pc = kzalloc(sizeof(struct page_cgroup), GFP_KERNEL);
> >> + if (pc == NULL)
> >> + goto err;
> >> +
> >> + rcu_read_lock();
> >> + /*
> >> + * We always charge the cgroup the mm_struct belongs to
> >> + * the mm_struct's mem_cgroup changes on task migration if the
> >> + * thread group leader migrates. It's possible that mm is not
> >> + * set, if so charge the init_mm (happens for pagecache usage).
> >> + */
> >> + if (!mm)
> >> + mm = &init_mm;
> >> +
> >> + mem = rcu_dereference(mm->mem_cgroup);
> >> + /*
> >> + * For every charge from the cgroup, increment reference
> >> + * count
> >> + */
> >> + css_get(&mem->css);
> >> + rcu_read_unlock();
> >> +
> >> + /*
> >> + * If we created the page_cgroup, we should free it on exceeding
> >> + * the cgroup limit.
> >> + */
> >> + if (res_counter_charge(&mem->res, 1)) {
> >> + css_put(&mem->css);
> >> + goto free_pc;
> >> + }
> >> +
> >> + lock_page_cgroup(page);
> >> + /*
> >> + * Check if somebody else beat us to allocating the page_cgroup
> >> + */
> >> + race_pc = page_get_page_cgroup(page);
> >> + if (race_pc) {
> >> + kfree(pc);
> >> + pc = race_pc;
> >> + atomic_inc(&pc->ref_cnt);
> >
> > This inc
> >
> >> + res_counter_uncharge(&mem->res, 1);
> >> + css_put(&mem->css);
> >> + goto done;
> >> + }
> >> +
> >> + atomic_set(&pc->ref_cnt, 1);
> >
> > combined with this set make me wonder...
> >
>
> I am not sure I understand this comment.
Is that inc needed? the pc is already associated with the page and
should thus already have a reference, so this inc would do 1->2, but we
then set it to 1 again. seems like a superfluous operation.
_______________________________________________
Containers mailing list
Containers@lists.linux-foundation.org
https://lists.linux-foundation.org/mailman/listinfo/containers
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| Re: [PATCH 01/29] task containersv11 basic task container framework [message #21003 is a reply to message #20999] |
Sun, 30 September 2007 07:10 |
Paul Menage
Messages: 642
Registered: September 2006
|
Senior Member |
|
|
On 9/29/07, Paul Jackson <pj@sgi.com> wrote:
> Paul M:
>
> This patch doesn't build for me in the following case. If I apply the
> rest of the containersv11 patches, it builds, but if I happen to bisect
> into this set of patches having applied only:
These patches weren't the normal style of incremental patchset - they
were direct substitutions of the previous patches in Andrew's -mm
tree.
Andrew's approach when encountering patches that are, e.g. missing
includes on some architectures, is to add a *-fix.patch to the series;
in this case it looks as though the fix for the missing ia64 include
was added towards the end of the series rather than directly after the
patch that needed it. Hence the fact that you couldn't compile until
that later patch had been pushed.
Andrew, is it practical for you to collapse any *-fix.patch files for
cgroups into the appropriate patches that they fix? That would
simplify the patch set a bit.
>
> task-containersv11-basic-task-container-framework.patch
>
> while using sn2_defconfig (with CONFIG_CGROUPS=y), then a build for
> arch ia64 fails with:
>
>
> $ make kernel/cgroup.o
> CHK include/linux/version.h
> CHK include/linux/utsrelease.h
> CALL scripts/checksyscalls.sh
> <stdin>:1389:2: warning: #warning syscall revokeat not implemented
> <stdin>:1393:2: warning: #warning syscall frevoke not implemented
> CC kernel/cgroup.o
> kernel/cgroup.c: In function 'cgroup_new_inode':
> kernel/cgroup.c:227: error: variable 'cgroup_backing_dev_info' has initializer but incomplete type
> kernel/cgroup.c:228: error: unknown field 'capabilities' specified in initializer
> kernel/cgroup.c:228: error: 'BDI_CAP_NO_ACCT_DIRTY' undeclared (first use in this function)
> kernel/cgroup.c:228: error: (Each undeclared identifier is reported only once
> kernel/cgroup.c:228: error: for each function it appears in.)
> kernel/cgroup.c:228: error: 'BDI_CAP_NO_WRITEBACK' undeclared (first use in this function)
> kernel/cgroup.c:228: warning: excess elements in struct initializer
> kernel/cgroup.c:228: warning: (near initialization for 'cgroup_backing_dev_info')
> kernel/cgroup.c:227: error: storage size of 'cgroup_backing_dev_info' isn't known
> kernel/cgroup.c:227: warning: unused variable 'cgroup_backing_dev_info'
> make[1]: *** [kernel/cgroup.o] Error 1
> make: *** [kernel/cgroup.o] Error 2
>
>
> I haven't tested building other partial applications of the
> containersv11 patch set; any initial sequence of them should build.
This wasn't strictly the containersv11 patch set - it was a set of
patches from -mm that needed to be replaced in order to rename
containers to control groups. The initial subset of these patches was
the same as containersv11, but it also included a bunch of other
people's patches (containerstats, memory controller, etc).
>
> ... also ... too bad the names of these patches all have 'container' in
> them, not 'cgroup'. I realize how this came to be, due to changing
> container to cgroup after these patches were named, but it sure would
> be nice if the final patch set record that gets layed down in Linus's
> history showed the correct name of 'cgroup' in these patch names.
Agreed. Presumably that would be fairly straightforward for Andrew to
do directly on his quilt tree - I kept the names the same as the
previous patches to make it simpler to match up old to new.
Paul
_______________________________________________
Containers mailing list
Containers@lists.linux-foundation.org
https://lists.linux-foundation.org/mailman/listinfo/containers
|
|
|
|
|
|
Goto Forum:
Current Time: Fri Sep 12 20:02:00 GMT 2025
Total time taken to generate the page: 0.07364 seconds
|
OpenVZ Forum
Members
Search
Help
Register
Login
Home
