Hi,

This patch splits the cgroup memory subsystem into two parts.
One is for tracking pages to find out the owners. The other is
for controlling how much amount of memory should be assigned to
each cgroup.

With this patch, you can use the page tracking mechanism even if
the memory subsystem is off.

Signed-off-by: Hirokazu Takahashi <taka@valinux.co.jp>


diff -udpr linux-2.6.26-rc2.cg0/include/linux/memcontrol.h linux-2.6.26-rc2/include/linux/memcontrol.h
--- linux-2.6.26-rc2.cg0/include/linux/memcontrol.h	2008-05-16 19:03:11.000000000 +0900
+++ linux-2.6.26-rc2/include/linux/memcontrol.h	2008-05-16 19:49:51.000000000 +0900
@@ -20,12 +20,61 @@
 #ifndef _LINUX_MEMCONTROL_H
 #define _LINUX_MEMCONTROL_H
 
+#include <linux/rcupdate.h>
+#include <linux/mm.h>
+#include <linux/smp.h>
+#include <linux/bit_spinlock.h>
+
 struct mem_cgroup;
 struct page_cgroup;
 struct page;
 struct mm_struct;
 
+#ifdef CONFIG_CGROUP_PAGE
+/*
+ * 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 at least two
+ * byte aligned (based on comments from Nick Piggin).  But since
+ * bit_spin_lock doesn't actually set that lock bit in a non-debug
+ * uniprocessor kernel, we should avoid setting it here too.
+ */
+#define PAGE_CGROUP_LOCK_BIT 	0x0
+#if defined(CONFIG_SMP) || defined(CONFIG_DEBUG_SPINLOCK)
+#define PAGE_CGROUP_LOCK 	(1 << PAGE_CGROUP_LOCK_BIT)
+#else
+#define PAGE_CGROUP_LOCK	0x0
+#endif
+
+/*
+ * A page_cgroup page is associated with every page descriptor. The
+ * page_cgroup helps us identify information about the cgroup
+ */
+struct page_cgroup {
 #ifdef CONFIG_CGROUP_MEM_RES_CTLR
+	struct list_head lru;		/* per cgroup LRU list */
+	struct mem_cgroup *mem_cgroup;
+#endif /* CONFIG_CGROUP_MEM_RES_CTLR */
+	struct page *page;
+	int ref_cnt;			/* cached, mapped, migrating */
+	int flags;
+};
+#define PAGE_CGROUP_FLAG_CACHE	(0x1)	/* charged as cache */
+#define PAGE_CGROUP_FLAG_ACTIVE (0x2)	/* page is active in this cgroup */
+
+static inline void lock_page_cgroup(struct page *page)
+{
+	bit_spin_lock(PAGE_CGROUP_LOCK_BIT, &page->page_cgroup);
+}
+
+static inline int try_lock_page_cgroup(struct page *page)
+{
+	return bit_spin_trylock(PAGE_CGROUP_LOCK_BIT, &page->page_cgroup);
+}
+
+static inline void unlock_page_cgroup(struct page *page)
+{
+	bit_spin_unlock(PAGE_CGROUP_LOCK_BIT, &page->page_cgroup);
+}
 
 #define page_reset_bad_cgroup(page)	((page)->page_cgroup = 0)
 
@@ -35,44 +84,15 @@ extern int mem_cgroup_charge(struct page
 extern int mem_cgroup_cache_charge(struct page *page, struct mm_struct *mm,
 					gfp_t gfp_mask);
 extern void mem_cgroup_uncharge_page(struct page *page);
-extern void mem_cgroup_move_lists(struct page *page, 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);
-extern void mem_cgroup_out_of_memory(struct mem_cgroup *mem, gfp_t gfp_mask);
-int task_in_mem_cgroup(struct task_struct *task, const struct mem_cgroup *mem);
-
-extern struct mem_cgroup *mem_cgroup_from_task(struct task_struct *p);
-
-#define mm_match_cgroup(mm, cgroup)	\
-	((cgroup) == mem_cgroup_from_task((mm)->owner))
 
 extern int
 mem_cgroup_prepare_migration(struct page *page, struct page *newpage);
 extern void mem_cgroup_end_migration(struct page *page);
 extern int mem_cgroup_getref(struct page *page);
 
-/*
- * For memory reclaim.
- */
-extern int mem_cgroup_calc_mapped_ratio(struct mem_cgroup *mem);
-extern long mem_cgroup_reclaim_imbalance(struct mem_cgroup *mem);
-
-extern int mem_cgroup_get_reclaim_priority(struct mem_cgroup *mem);
-extern void mem_cgroup_note_reclaim_priority(struct mem_cgroup *mem,
-							int priority);
-extern void mem_cgroup_record_reclaim_priority(struct mem_cgroup *mem,
-							int priority);
-
-extern long mem_cgroup_calc_reclaim_active(struct mem_cgroup *mem,
-				struct zone *zone, int priority);
-extern long mem_cgroup_calc_reclaim_inactive(struct mem_cgroup *mem,
-				struct zone *zone, int priority);
+extern void page_cgroup_init(void);
 
-#else /* CONFIG_CGROUP_MEM_RES_CTLR */
+#else /* CONFIG_CGROUP_PAGE */
 static inline void page_reset_bad_cgroup(struct page *page)
 {
 }
@@ -98,33 +118,70 @@ static inline void mem_cgroup_uncharge_p
 {
 }
 
-static inline void mem_cgroup_move_lists(struct page *page, bool active)
+static inline int mem_cgroup_prepare_migration(struct page *page)
 {
+	return 0;
 }
 
-static inline int mm_match_cgroup(struct mm_struct *mm, struct mem_cgroup *mem)
+static inline void mem_cgroup_end_migration(struct page *page)
 {
-	return 1;
 }
 
-static inline int task_in_mem_cgroup(struct task_struct *task,
-				     const struct mem_cgroup *mem)
+static inline void mem_cgroup_getref(struct page *page)
 {
-	return 1;
 }
+#endif /* CONFIG_CGROUP_PAGE */
 
-static inline int
-mem_cgroup_prepare_migration(struct page *page, struct page *newpage)
+
+#ifdef CONFIG_CGROUP_MEM_RES_CTLR
+
+extern void mem_cgroup_move_lists(struct page *page, 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);
+extern void mem_cgroup_out_of_memory(struct mem_cgroup *mem, gfp_t gfp_mask);
+int task_in_mem_cgroup(struct task_struct *task, const struct mem_cgroup *mem);
+
+extern struct mem_cgroup *mem_cgroup_from_task(struct task_struct *p);
+
+#define mm_match_cgroup(mm, cgroup)	\
+	((cgroup) == mem_cgroup_from_task((mm)->owner))
+
+/*
+ * For memory reclaim.
+ */
+extern int mem_cgroup_calc_mapped_ratio(struct mem_cgroup *mem);
+extern long mem_cgroup_reclaim_imbalance(struct mem_cgroup *mem);
+
+extern int mem_cgroup_get_reclaim_priority(struct mem_cgroup *mem);
+extern void mem_cgroup_note_reclaim_priority(struct mem_cgroup *mem,
+							int priority);
+extern void mem_cgroup_record_reclaim_priority(struct mem_cgroup *mem,
+							int priority);
+
+extern long mem_cgroup_calc_reclaim_active(struct mem_cgroup *mem,
+				struct zone *zone, int priority);
+extern long mem_cgroup_calc_reclaim_inactive(struct mem_cgroup *mem,
+				struct zone *zone, int priority);
+
+#else /* CONFIG_CGROUP_MEM_RES_CTLR */
+
+static inline void mem_cgroup_move_lists(struct page *page, bool active)
 {
-	return 0;
 }
 
-static inline void mem_cgroup_end_migration(struct page *page)
+static inline int mm_match_cgroup(struct mm_struct *mm, struct mem_cgroup *mem)
 {
+	return 1;
 }
 
-static inline void mem_cgroup_getref(struct page *page)
+static inline int task_in_mem_cgroup(struct task_struct *task,
+				     const struct mem_cgroup *mem)
 {
+	return 1;
 }
 
 static inline int mem_cgroup_calc_mapped_ratio(struct mem_cgroup *mem)
@@ -163,7 +220,7 @@ static inline long mem_cgroup_calc_recla
 {
 	return 0;
 }
-#endif /* CONFIG_CGROUP_MEM_CONT */
+#endif /* CONFIG_CGROUP_MEM_RES_CTLR */
 
 #endif /* _LINUX_MEMCONTROL_H */
 
diff -udpr linux-2.6.26-rc2.cg0/include/linux/mm_types.h linux-2.6.26-rc2/include/linux/mm_types.h
--- linux-2.6.26-rc2.cg0/include/linux/mm_types.h	2008-05-16 19:03:11.000000000 +0900
+++ linux-2.6.26-rc2/include/linux/mm_types.h	2008-05-16 19:03:43.000000000 +0900
@@ -91,7 +91,7 @@ struct page {
 	void *virtual;			/* Kernel virtual address (NULL if
 					   not kmapped, ie. highmem) */
 #endif /* WANT_PAGE_VIRTUAL */
-#ifdef CONFIG_CGROUP_MEM_RES_CTLR
+#ifdef CONFIG_CGROUP_PAGE
 	unsigned long page_cgroup;
 #endif
 #ifdef CONFIG_PAGE_OWNER
diff -udpr linux-2.6.26-rc2.cg0/init/Kconfig linux-2.6.26-rc2/init/Kconfig
--- linux-2.6.26-rc2.cg0/init/Kconfig	2008-05-16 19:03:11.000000000 +0900
+++ linux-2.6.26-rc2/init/Kconfig	2008-05-16 19:03:43.000000000 +0900
@@ -407,6 +407,10 @@ config CGROUP_MEM_RES_CTLR
 	  This config option also selects MM_OWNER config option, which
 	  could in turn add some fork/exit overhead.
 
+config CGROUP_PAGE
+	def_bool y
+	depends on CGROUP_MEM_RES_CTLR
+
 config SYSFS_DEPRECATED
 	bool
 
diff -udpr linux-2.6.26-rc2.cg0/mm/Makefile linux-2.6.26-rc2/mm/Makefile
--- linux-2.6.26-rc2.cg0/mm/Makefile	2008-05-16 19:03:11.000000000 +0900
+++ linux-2.6.26-rc2/mm/Makefile	2008-05-16 19:03:43.000000000 +0900
@@ -33,5 +33,5 @@ obj-$(CONFIG_FS_XIP) += filemap_xip.o
 obj-$(CONFIG_MIGRATION) += migrate.o
 obj-$(CONFIG_SMP) += allocpercpu.o
 obj-$(CONFIG_QUICKLIST) += quicklist.o
-obj-$(CONFIG_CGROUP_MEM_RES_CTLR) += memcontrol.o
+obj-$(CONFIG_CGROUP_PAGE) += memcontrol.o
 
diff -udpr linux-2.6.26-rc2.cg0/mm/memcontrol.c linux-2.6.26-rc2/mm/memcontrol.c
--- linux-2.6.26-rc2.cg0/mm/memcontrol.c	2008-05-16 19:03:11.000000000 +0900
+++ linux-2.6.26-rc2/mm/memcontrol.c	2008-05-19 11:39:43.000000000 +0900
@@ -35,10 +35,17 @@
 
 #include <asm/uaccess.h>
 
-struct cgroup_subsys mem_cgroup_subsys __read_mostly;
 static struct kmem_cache *page_cgroup_cache __read_mostly;
+
+#ifdef CONFIG_CGROUP_MEM_RES_CTLR
+struct cgroup_subsys mem_cgroup_subsys __read_mostly;
 #define MEM_CGROUP_RECLAIM_RETRIES	5
 
+static inline int mem_cgroup_disabled(void)
+{
+	return mem_cgroup_subsys.disabled;
+}
+
 /*
  * Statistics for memory cgroup.
  */
@@ -144,34 +151,6 @@ struct mem_cgroup {
 };
 static struct mem_cgroup init_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 at least two
- * byte aligned (based on comments from Nick Piggin).  But since
- * bit_spin_lock doesn't actually set that lock bit in a non-debug
- * uniprocessor kernel, we should avoid setting it here too.
- */
-#define PAGE_CGROUP_LOCK_BIT 	0x0
-#if defined(CONFIG_SMP) || defined(CONFIG_DEBUG_SPINLOCK)
-#define PAGE_CGROUP_LOCK 	(1 << PAGE_CGROUP_LOCK_BIT)
-#else
-#define PAGE_CGROUP_LOCK	0x0
-#endif
-
-/*
- * A page_cgroup