Memory controller: cgroups setup
Balbir Singh [Thu, 7 Feb 2008 08:13:50 +0000 (00:13 -0800)]
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>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

include/linux/cgroup_subsys.h
include/linux/memcontrol.h [new file with mode: 0644]
init/Kconfig
mm/Makefile
mm/memcontrol.c [new file with mode: 0644]

index 9ec4318..228235c 100644 (file)
@@ -37,3 +37,8 @@ SUBSYS(cpuacct)
 
 /* */
 
+#ifdef CONFIG_CGROUP_MEM_CONT
+SUBSYS(mem_cgroup)
+#endif
+
+/* */
diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h
new file mode 100644 (file)
index 0000000..3f121b2
--- /dev/null
@@ -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 */
+
index d372bd6..95ac265 100644 (file)
@@ -397,6 +397,13 @@ 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
index 4af5dff..9f117ba 100644 (file)
@@ -32,4 +32,5 @@ 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 --git a/mm/memcontrol.c b/mm/memcontrol.c
new file mode 100644 (file)
index 0000000..80e48cd
--- /dev/null
@@ -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,
+};