Add an owner to the mm_struct (v4)

diff -puN include/linux/mm_types.h~memory-controller-add-mm-owner include/linux/mm_types.h
@@ -227,8 +227,9 @@ struct mm_struct {
 	/* aio bits */
 	rwlock_t		ioctx_list_lock;
 	struct kioctx		*ioctx_list;
-#ifdef CONFIG_CGROUP_MEM_RES_CTLR
-	struct mem_cgroup *mem_cgroup;
+#ifdef CONFIG_MM_OWNER
+	struct task_struct *owner;	/* The thread group leader that */
+					/* owns the mm_struct.		*/
 #endif
 
 #ifdef CONFIG_PROC_FS

diff -puN kernel/fork.c~memory-controller-add-mm-owner kernel/fork.c
@@ -358,14 +358,13 @@ 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);
+	mm_init_owner(mm, p);
 
 	if (likely(!mm_alloc_pgd(mm))) {
 		mm->def_flags = 0;
 		return mm;
 	}
 
-	mm_free_cgroup(mm);
 	free_mm(mm);
 	return NULL;
 }

@@ -394,7 +393,6 @@ void __mmdrop(struct mm_struct *mm)
 {
 	BUG_ON(mm == &init_mm);
 	mm_free_pgd(mm);
-	mm_free_cgroup(mm);
 	destroy_context(mm);
 	free_mm(mm);
 }

@@ -995,6 +993,13 @@ static void rt_mutex_init_task(struct ta
 #endif
 }
 
+#ifdef CONFIG_MM_OWNER
+void mm_init_owner(struct mm_struct *mm, struct task_struct *p)
+{
+	mm->owner = p;
+}
+#endif /* CONFIG_MM_OWNER */
+
 /*
  * This creates a new process as a copy of the old one,
  * but does not actually start it yet.

diff -puN include/linux/memcontrol.h~memory-controller-add-mm-owner include/linux/memcontrol.h
@@ -27,9 +27,6 @@ struct mm_struct;
 
 #ifdef CONFIG_CGROUP_MEM_RES_CTLR
 
-extern void mm_init_cgroup(struct mm_struct *mm, struct task_struct *p);
-extern void mm_free_cgroup(struct mm_struct *mm);
-
 #define page_reset_bad_cgroup(page)	((page)->page_cgroup = 0)
 
 extern struct page_cgroup *page_get_page_cgroup(struct page *page);

@@ -48,8 +45,10 @@ extern unsigned long mem_cgroup_isolate_
 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) == rcu_dereference((mm)->mem_cgroup))
+	((cgroup) == mem_cgroup_from_task((mm)->owner))
 
 extern int mem_cgroup_prepare_migration(struct page *page);
 extern void mem_cgroup_end_migration(struct page *page);

@@ -73,15 +72,6 @@ extern long mem_cgroup_calc_reclaim_inac
 				struct zone *zone, int priority);
 
 #else /* CONFIG_CGROUP_MEM_RES_CTLR */
-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_reset_bad_cgroup(struct page *page)
 {
 }

@@ -172,6 +162,7 @@ static inline long mem_cgroup_calc_recla
 {
 	return 0;
 }
+
 #endif /* CONFIG_CGROUP_MEM_CONT */
 
 #endif /* _LINUX_MEMCONTROL_H */

diff -puN mm/memcontrol.c~memory-controller-add-mm-owner mm/memcontrol.c
@@ -238,26 +238,12 @@ static struct mem_cgroup *mem_cgroup_fro
 				css);
 }
 
-static struct mem_cgroup *mem_cgroup_from_task(struct task_struct *p)
+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);
-}
-
 static inline int page_cgroup_locked(struct page *page)
 {
 	return bit_spin_is_locked(PAGE_CGROUP_LOCK_BIT, &page->page_cgroup);

@@ -478,6 +464,7 @@ unsigned long mem_cgroup_isolate_pages(u
 	int zid = zone_idx(z);
 	struct mem_cgroup_per_zone *mz;
 
+	BUG_ON(!mem_cont);
 	mz = mem_cgroup_zoneinfo(mem_cont, nid, zid);
 	if (active)
 		src = &mz->active_list;

@@ -576,7 +563,7 @@ retry:
 		mm = &init_mm;
 
 	rcu_read_lock();
-	mem = rcu_dereference(mm->mem_cgroup);
+	mem = mem_cgroup_from_task(rcu_dereference(mm->owner));
 	/*
 	 * For every charge from the cgroup, increment reference count
 	 */

@@ -990,7 +977,6 @@ mem_cgroup_create(struct cgroup_subsys *
 
 	if (unlikely((cont->parent) == NULL)) {
 		mem = &init_mem_cgroup;
-		init_mm.mem_cgroup = mem;
 		page_cgroup_cache = KMEM_CACHE(page_cgroup, SLAB_PANIC);
 	} else
 		mem = kzalloc(sizeof(struct mem_cgroup), GFP_KERNEL);

@@ -1072,7 +1058,6 @@ static void mem_cgroup_move_task(struct 
 		goto out;
 
 	css_get(&mem->css);
-	rcu_assign_pointer(mm->mem_cgroup, mem);
 	css_put(&old_mem->css);
 
 out:

diff -puN include/linux/sched.h~memory-controller-add-mm-owner include/linux/sched.h
@@ -2130,6 +2130,20 @@ static inline void migration_init(void)
 
 #define TASK_STATE_TO_CHAR_STR "RSDTtZX"
 
+#ifdef CONFIG_MM_OWNER
+extern void mm_update_next_owner(struct mm_struct *mm);
+extern void mm_init_owner(struct mm_struct *mm, struct task_struct *p);
+#else
+static inline void
+mm_update_next_owner(struct mm_struct *mm, struct task_struct *p)
+{
+}
+
+static inline void mm_init_owner(struct mm_struct *mm, struct task_struct *p)
+{
+}
+#endif /* CONFIG_MM_OWNER */
+
 #endif /* __KERNEL__ */
 
 #endif

78 lines not on this page · raw message

diff -puN include/linux/init_task.h~memory-controller-add-mm-owner include/linux/init_task.h
@@ -57,6 +57,7 @@
 	.page_table_lock =  __SPIN_LOCK_UNLOCKED(name.page_table_lock),	\
 	.mmlist		= LIST_HEAD_INIT(name.mmlist),		\
 	.cpu_vm_mask	= CPU_MASK_ALL,				\
+	.owner		= &init_task,				\
 }
 
 #define INIT_SIGNALS(sig) {						\

@@ -199,7 +200,6 @@ extern struct group_info init_groups;
 	INIT_LOCKDEP							\
 }
 
-
 #define INIT_CPU_TIMERS(cpu_timers)					\
 {									\
 	LIST_HEAD_INIT(cpu_timers[0]),					\

diff -puN fs/exec.c~memory-controller-add-mm-owner fs/exec.c
@@ -735,6 +735,7 @@ static int exec_mmap(struct mm_struct *m
 	tsk->active_mm = mm;
 	activate_mm(active_mm, mm);
 	task_unlock(tsk);
+	mm_update_next_owner(mm);
 	arch_pick_mmap_layout(mm);
 	if (old_mm) {
 		up_read(&old_mm->mmap_sem);

diff -puN kernel/cgroup.c~memory-controller-add-mm-owner kernel/cgroup.c
@@ -118,6 +118,7 @@ static int root_count;
  * be called.
  */
 static int need_forkexit_callback;
+static int need_mm_owner_callback;
 
 /* convenient tests for these bits */
 inline int cgroup_is_removed(const struct cgroup *cgrp)

@@ -2481,6 +2482,7 @@ static void __init cgroup_init_subsys(st
 	}
 
 	need_forkexit_callback |= ss->fork || ss->exit;
+	need_mm_owner_callback |= !!ss->mm_owner_changed;
 
 	ss->active = 1;
 }

@@ -2717,6 +2719,28 @@ void cgroup_fork_callbacks(struct task_s
 	}
 }
 
+#ifdef CONFIG_MM_OWNER
+/**
+ * cgroup_mm_owner_callbacks - run callbacks when the mm->owner changes
+ * @p: the new owner
+ *
+ * Called on every change to mm->owner. mm_init_owner() does not
+ * invoke this routine, since it assigns the mm->owner the first time
+ * and does not change it.
+ */
+void cgroup_mm_owner_callbacks(struct task_struct *old, struct task_struct *new)
+{
+	if (need_mm_owner_callback) {
+		int i;
+		for (i = 0; i < CGROUP_SUBSYS_COUNT; i++) {
+			struct cgroup_subsys *ss = subsys[i];
+			if (ss->mm_owner_changed)
+				ss->mm_owner_changed(ss, old, new);
+		}
+	}
+}
+#endif /* CONFIG_MM_OWNER */
+
 /**
  * cgroup_post_fork - called on a new task after adding it to the task list
  * @child: the task in question

diff -puN include/linux/cgroup.h~memory-controller-add-mm-owner include/linux/cgroup.h
@@ -292,6 +292,12 @@ struct cgroup_subsys {
 			struct cgroup *cgrp);
 	void (*post_clone)(struct cgroup_subsys *ss, struct cgroup *cgrp);
 	void (*bind)(struct cgroup_subsys *ss, struct cgroup *root);
+	/*
+	 * This routine is called with the task_lock of mm->owner held
+	 */
+	void (*mm_owner_changed)(struct cgroup_subsys *ss,
+					struct task_struct *old,
+					struct task_struct *new);
 	int subsys_id;
 	int active;
 	int disabled;

@@ -377,4 +383,13 @@ static inline int cgroupstats_build(stru
 
 #endif /* !CONFIG_CGROUPS */
 
+#ifdef CONFIG_MM_OWNER
+extern void
+cgroup_mm_owner_callbacks(struct task_struct *old, struct task_struct *new);
+#else /* !CONFIG_MM_OWNER */
+static inline void
+cgroup_mm_owner_callbacks(struct task_struct *old, struct task_struct *new)
+{
+}
+#endif /* CONFIG_MM_OWNER */
 #endif /* _LINUX_CGROUP_H */