cgroup源码分析3——cgroup层级的mount流程

1
2

# mount -t cgroup -o cpu,cpuacct   none  /sys/fs/cgroup/cpu,cpuacct
# mount -t cgroup -o pids none /sysfs/cgroup/cpu,cpuacct

  1
  2
  3
  4
  5
  6
  7
  8
  9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154

static struct dentry *cgroup_mount(struct file_system_type *fs_type,
			 int flags, const char *unused_dev_name,
			 void *data)
{
	struct cgroup_sb_opts opts;
	struct cgroupfs_root *root;
	int ret = 0;
	struct super_block *sb;
	struct cgroupfs_root *new_root;
	struct inode *inode;

	/* First find the desired set of subsystems */
	mutex_lock(&cgroup_mutex);
	ret = parse_cgroupfs_options(data, &opts);
	mutex_unlock(&cgroup_mutex);
	if (ret)
		goto out_err;

	/*
	 * Allocate a new cgroup root. We may not need it if we're
	 * reusing an existing hierarchy.
	 */
	new_root = cgroup_root_from_opts(&opts); // 解析mount时的options选项
	if (IS_ERR(new_root)) {
		ret = PTR_ERR(new_root);
		goto drop_modules;
	}
	opts.new_root = new_root;

	/* Locate an existing or new sb for this hierarchy */// 分配一个新的cgroupfs_root
	sb = sget(fs_type, cgroup_test_super, cgroup_set_super, 0, &opts);
	if (IS_ERR(sb)) {
		ret = PTR_ERR(sb);
		cgroup_drop_root(opts.new_root);
		goto drop_modules;
	}

	root = sb->s_fs_info;
	BUG_ON(!root);
	if (root == opts.new_root) { // 新的挂载，说明这是一个新的层级
		/* We used the new root structure, so this is a new hierarchy */
		struct list_head tmp_cg_links;
		struct cgroup *root_cgrp = &root->top_cgroup;
		struct cgroupfs_root *existing_root;
		const struct cred *cred;
		int i;
		struct css_set *cg;

		BUG_ON(sb->s_root != NULL);
		// 获取挂载点的inode
		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);
		mutex_lock(&cgroup_root_mutex);

		/* Check for name clashes with existing mounts */
		ret = -EBUSY;
		if (strlen(root->name))
			for_each_active_root(existing_root)
				if (!strcmp(existing_root->name, root->name))
					goto unlock_drop;

		/*
		 * We're accessing css_set_count without locking
		 * css_set_lock here, but that's OK - it can only be
		 * increased by someone holding cgroup_lock, and
		 * that's us. The worst that can happen is that we
		 * have some link structures left over
		 */// 分配css_set_count个cg_cgroup_link结构
		ret = allocate_cg_links(css_set_count, &tmp_cg_links);
		if (ret)
			goto unlock_drop;

		ret = rebind_subsystems(root, root->subsys_mask);
		if (ret == -EBUSY) {
			free_cg_links(&tmp_cg_links);
			goto unlock_drop;
		}
		/*
		 * There must be no failure case after here, since rebinding
		 * takes care of subsystems' refcounts, which are explicitly
		 * dropped in the failure exit path.
		 */

		/* EBUSY should be the only error here */
		BUG_ON(ret);
		// 将该cgroupfs_root添加到链表roots中
		list_add(&root->root_list, &roots);
		root_count++;

		sb->s_root->d_fsdata = root_cgrp;
		root->top_cgroup.dentry = sb->s_root;

		/* Link the top cgroup in this hierarchy into all
		 * the css_set objects */
		write_lock(&css_set_lock);
		hash_for_each(css_set_table, i, cg, hlist)
			link_css_set(&tmp_cg_links, cg, root_cgrp);
		write_unlock(&css_set_lock);

		free_cg_links(&tmp_cg_links);

		BUG_ON(!list_empty(&root_cgrp->children));
		BUG_ON(root->number_of_cgroups != 1);

		cred = override_creds(&init_cred);
		// 创建对应的cgroup控制文件
		cgroup_populate_dir(root_cgrp, true, root->subsys_mask);
		revert_creds(cred);
		mutex_unlock(&cgroup_root_mutex);
		mutex_unlock(&cgroup_mutex);
		mutex_unlock(&inode->i_mutex);
	} else { // 已经挂载了，不需要做什么
		/*
		 * We re-used an existing hierarchy - the new root (if
		 * any) is not needed
		 */
		cgroup_drop_root(opts.new_root);

		if (root->flags != opts.flags) {
			if ((root->flags | opts.flags) & CGRP_ROOT_SANE_BEHAVIOR) {
				pr_err("cgroup: sane_behavior: new mount options should match the existing superblock\n");
				ret = -EINVAL;
				goto drop_new_super;
			} else {
				pr_warning("cgroup: new mount options do not match the existing superblock, will be ignored\n");
			}
		}

		/* no subsys rebinding, so refcounts don't change */
		drop_parsed_module_refcounts(opts.subsys_mask);
	}

	kfree(opts.release_agent);
	kfree(opts.name);
	return dget(sb->s_root);

 unlock_drop:
	mutex_unlock(&cgroup_root_mutex);
	mutex_unlock(&cgroup_mutex);
	mutex_unlock(&inode->i_mutex);
 drop_new_super:
	deactivate_locked_super(sb);
 drop_modules:
	drop_parsed_module_refcounts(opts.subsys_mask);
 out_err:
	kfree(opts.release_agent);
	kfree(opts.name);
	return ERR_PTR(ret);
}

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28

static struct cgroupfs_root *cgroup_root_from_opts(struct cgroup_sb_opts *opts)
{
	struct cgroupfs_root *root;

	if (!opts->subsys_mask && !opts->none)
		return NULL;

	root = kzalloc(sizeof(*root), GFP_KERNEL);
	if (!root)
		return ERR_PTR(-ENOMEM);

	if (!init_root_id(root)) { // 重点
		kfree(root);
		return ERR_PTR(-ENOMEM);
	}
	init_cgroup_root(root);

	root->subsys_mask = opts->subsys_mask;
	root->flags = opts->flags;
	ida_init(&root->cgroup_ida);
	if (opts->release_agent)
		strcpy(root->release_agent_path, opts->release_agent);
	if (opts->name)
		strcpy(root->name, opts->name);
	if (opts->cpuset_clone_children)
		set_bit(CGRP_CPUSET_CLONE_CHILDREN, &root->top_cgroup.flags);
	return root;
}

1
2
3

static DEFINE_IDA(hierarchy_ida);
static int next_hierarchy_id;
static DEFINE_SPINLOCK(hierarchy_id_lock);

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24

static bool init_root_id(struct cgroupfs_root *root)
{
	int ret = 0;

	do {
		if (!ida_pre_get(&hierarchy_ida, GFP_KERNEL))
			return false;
		spin_lock(&hierarchy_id_lock);
		/* Try to allocate the next unused ID */
		ret = ida_get_new_above(&hierarchy_ida, next_hierarchy_id,
					&root->hierarchy_id);
		if (ret == -ENOSPC)
			/* Try again starting from 0 */
			ret = ida_get_new(&hierarchy_ida, &root->hierarchy_id);
		if (!ret) {
			next_hierarchy_id = root->hierarchy_id + 1;
		} else if (ret != -EAGAIN) {
			/* Can only get here if the 31-bit IDR is full ... */
			BUG_ON(ret);
		}
		spin_unlock(&hierarchy_id_lock);
	} while (ret);
	return true;
}

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19

static int cgroup_test_super(struct super_block *sb, void *data)
{
	struct cgroup_sb_opts *opts = data;
	struct cgroupfs_root *root = sb->s_fs_info;

	/* If we asked for a name then it must match */
	if (opts->name && strcmp(opts->name, root->name))
		return 0;

	/*
	 * If we asked for subsystems (or explicitly for no
	 * subsystems) then they must match
	 */
	if ((opts->subsys_mask || opts->none)
	    && (opts->subsys_mask != root->subsys_mask))
		return 0;

	return 1;
}

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

static int cgroup_set_super(struct super_block *sb, void *data)
{
	int ret;
	struct cgroup_sb_opts *opts = data;

	/* If we don't have a new root, we can't set up a new sb */
	if (!opts->new_root)
		return -EINVAL;

	BUG_ON(!opts->subsys_mask && !opts->none);

	ret = set_anon_super(sb, NULL);
	if (ret)
		return ret;

	sb->s_fs_info = opts->new_root;
	opts->new_root->sb = sb;

	sb->s_blocksize = PAGE_CACHE_SIZE;
	sb->s_blocksize_bits = PAGE_CACHE_SHIFT;
	sb->s_magic = CGROUP_SUPER_MAGIC;
	sb->s_op = &cgroup_ops;

	return 0;
}

1
2
3
4
5
6

static const struct super_operations cgroup_ops = {
	.statfs = simple_statfs,
	.drop_inode = generic_delete_inode,
	.show_options = cgroup_show_options,
	.remount_fs = cgroup_remount,
};

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93

/*
 * Call with cgroup_mutex held. Drops reference counts on modules, including
 * any duplicate ones that parse_cgroupfs_options took. If this function
 * returns an error, no reference counts are touched.
 */
static int rebind_subsystems(struct cgroupfs_root *root,
			      unsigned long final_subsys_mask)
{
	unsigned long added_mask, removed_mask;
	struct cgroup *cgrp = &root->top_cgroup;
	int i;

	BUG_ON(!mutex_is_locked(&cgroup_mutex));
	BUG_ON(!mutex_is_locked(&cgroup_root_mutex));

	removed_mask = root->actual_subsys_mask & ~final_subsys_mask;
	added_mask = final_subsys_mask & ~root->actual_subsys_mask;
	/* Check that any added subsystems are currently free */
	for (i = 0; i < CGROUP_SUBSYS_COUNT; i++) {
		unsigned long bit = 1UL << i;
		struct cgroup_subsys *ss = subsys[i];
		if (!(bit & added_mask))
			continue;
		/*
		 * Nobody should tell us to do a subsys that doesn't exist:
		 * parse_cgroupfs_options should catch that case and refcounts
		 * ensure that subsystems won't disappear once selected.
		 */
		BUG_ON(ss == NULL);
		if (ss->root != &rootnode) {
			/* Subsystem isn't free */
			return -EBUSY;
		}
	}

	/* Currently we don't handle adding/removing subsystems when
	 * any child cgroups exist. This is theoretically supportable
	 * but involves complex error handling, so it's being left until
	 * later */
	if (root->number_of_cgroups > 1)
		return -EBUSY;

	/* Process each subsystem */
	for (i = 0; i < CGROUP_SUBSYS_COUNT; i++) {
		struct cgroup_subsys *ss = subsys[i];
		unsigned long bit = 1UL << i;
		if (bit & added_mask) {
			/* We're binding this subsystem to this hierarchy */
			BUG_ON(ss == NULL);
			BUG_ON(cgrp->subsys[i]);
			BUG_ON(!dummytop->subsys[i]);
			BUG_ON(dummytop->subsys[i]->cgroup != dummytop);
			cgrp->subsys[i] = dummytop->subsys[i];
			cgrp->subsys[i]->cgroup = cgrp;
			list_move(&ss->sibling, &root->subsys_list);
			ss->root = root;
			if (ss->bind)
				ss->bind(cgrp);
			/* refcount was already taken, and we're keeping it */
		} else if (bit & removed_mask) {
			/* We're removing this subsystem */
			BUG_ON(ss == NULL);
			BUG_ON(cgrp->subsys[i] != dummytop->subsys[i]);
			BUG_ON(cgrp->subsys[i]->cgroup != cgrp);
			if (ss->bind)
				ss->bind(dummytop);
			dummytop->subsys[i]->cgroup = dummytop;
			cgrp->subsys[i] = NULL;
			subsys[i]->root = &rootnode;
			list_move(&ss->sibling, &rootnode.subsys_list);
			/* subsystem is now free - drop reference on module */
			module_put(ss->module);
		} else if (bit & final_subsys_mask) {
			/* Subsystem state should already exist */
			BUG_ON(ss == NULL);
			BUG_ON(!cgrp->subsys[i]);
			/*
			 * a refcount was taken, but we already had one, so
			 * drop the extra reference.
			 */
			module_put(ss->module);
#ifdef CONFIG_MODULE_UNLOAD
			BUG_ON(ss->module && !module_refcount(ss->module));
#endif
		} else {
			/* Subsystem state shouldn't exist */
			BUG_ON(cgrp->subsys[i]);
		}
	}
	root->subsys_mask = root->actual_subsys_mask = final_subsys_mask;

	return 0;
}

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42

/**
 * cgroup_populate_dir - selectively creation of files in a directory
 * @cgrp: target cgroup
 * @base_files: true if the base files should be added
 * @subsys_mask: mask of the subsystem ids whose files should be added
 */
static int cgroup_populate_dir(struct cgroup *cgrp, bool base_files,
			       unsigned long subsys_mask)
{
	int err;
	struct cgroup_subsys *ss;

	if (base_files) { // files 定义了cgroup的基本文件, true代表添加文件
		err = cgroup_addrm_files(cgrp, NULL, files, true);
		if (err < 0)
			return err;
	}

	/* process cftsets of each subsystem */
	for_each_subsys(cgrp->root, ss) { // 对于该层级上挂载的每一个cgroup子系统，创建控制文件
		struct cftype_set *set;
		if (!test_bit(ss->subsys_id, &subsys_mask))
			continue;

		list_for_each_entry(set, &ss->cftsets, node)
			cgroup_addrm_files(cgrp, ss, set->cfts, true);
	}

	/* This cgroup is ready now */
	for_each_subsys(cgrp->root, ss) {
		struct cgroup_subsys_state *css = cgrp->subsys[ss->subsys_id];
		/*
		 * Update id->css pointer and make this css visible from
		 * CSS ID functions. This pointer will be dereferened
		 * from RCU-read-side without locks.
		 */
		if (css->id)
			rcu_assign_pointer(css->id->css, css);
	}

	return 0;
}

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50

/*
 * for the common functions, 'private' gives the type of file
 */
/* for hysterical raisins, we can't put this on the older files */
#define CGROUP_FILE_GENERIC_PREFIX "cgroup."
static struct cftype files[] = {
	{
		.name = "tasks",
		.open = cgroup_tasks_open,
		.write_u64 = cgroup_tasks_write,
		.release = cgroup_pidlist_release,
		.mode = S_IRUGO | S_IWUSR,
	},
	{
		.name = CGROUP_FILE_GENERIC_PREFIX "procs",
		.open = cgroup_procs_open,
		.write_u64 = cgroup_procs_write,
		.release = cgroup_pidlist_release,
		.mode = S_IRUGO | S_IWUSR,
	},
	{
		.name = "notify_on_release",
		.read_u64 = cgroup_read_notify_on_release,
		.write_u64 = cgroup_write_notify_on_release,
	},
	{
		.name = CGROUP_FILE_GENERIC_PREFIX "event_control",
		.write_string = cgroup_write_event_control,
		.mode = S_IWUGO,
	},
	{
		.name = "cgroup.clone_children",
		.flags = CFTYPE_INSANE,
		.read_u64 = cgroup_clone_children_read,
		.write_u64 = cgroup_clone_children_write,
	},
	{
		.name = "cgroup.sane_behavior",
		.flags = CFTYPE_ONLY_ON_ROOT,
		.read_seq_string = cgroup_sane_behavior_show,
	},
	{
		.name = "release_agent",
		.flags = CFTYPE_ONLY_ON_ROOT,
		.read_seq_string = cgroup_release_agent_show,
		.write_string = cgroup_release_agent_write,
		.max_write_len = PATH_MAX,
	},
	{ }	/* terminate */
};

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67

static int cgroup_remount(struct super_block *sb, int *flags, char *data)
{
	int ret = 0;
	struct cgroupfs_root *root = sb->s_fs_info;
	struct cgroup *cgrp = &root->top_cgroup;
	struct cgroup_sb_opts opts;
	unsigned long added_mask, removed_mask;
	// __DEVEL__sane_behavior 指定后，不允许进行remount操作
	if (root->flags & CGRP_ROOT_SANE_BEHAVIOR) {
		pr_err("cgroup: sane_behavior: remount is not allowed\n");
		return -EINVAL;
	}

	mutex_lock(&cgrp->dentry->d_inode->i_mutex);
	mutex_lock(&cgroup_mutex);
	mutex_lock(&cgroup_root_mutex);

	/* See what subsystems are wanted */ // 解析remount的选项
	ret = parse_cgroupfs_options(data, &opts);
	if (ret)
		goto out_unlock;
	// 不建议remount时修改层级的子系统或者其他选项，该功能已经废弃
	if (opts.subsys_mask != root->actual_subsys_mask || opts.release_agent)
		pr_warning("cgroup: option changes via remount are deprecated (pid=%d comm=%s)\n",
			   task_tgid_nr(current), current->comm);

	added_mask = opts.subsys_mask & ~root->subsys_mask;
	removed_mask = root->subsys_mask & ~opts.subsys_mask;

	// flags和name在remount时不允许改变
	/* Don't allow flags or name to change at remount */
	if (opts.flags != root->flags ||
	    (opts.name && strcmp(opts.name, root->name))) {
		ret = -EINVAL;
		drop_parsed_module_refcounts(opts.subsys_mask);
		goto out_unlock;
	}

	/*
	 * Clear out the files of subsystems that should be removed, do
	 * this before rebind_subsystems, since rebind_subsystems may
	 * change this hierarchy's subsys_list.
	 *///删除 要删除的cgroup子系统的控制文件
	cgroup_clear_directory(cgrp->dentry, false, removed_mask);

	ret = rebind_subsystems(root, opts.subsys_mask);
	if (ret) {
		/* rebind_subsystems failed, re-populate the removed files */
		cgroup_populate_dir(cgrp, false, removed_mask);
		drop_parsed_module_refcounts(opts.subsys_mask);
		goto out_unlock;
	}

	/* re-populate subsystem files */
	// 添加 要添加的cgroup子系统的控制文件
	cgroup_populate_dir(cgrp, false, added_mask);

	if (opts.release_agent)
		strcpy(root->release_agent_path, opts.release_agent);
 out_unlock:
	kfree(opts.release_agent);
	kfree(opts.name);
	mutex_unlock(&cgroup_root_mutex);
	mutex_unlock(&cgroup_mutex);
	mutex_unlock(&cgrp->dentry->d_inode->i_mutex);
	return ret;
}

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49

static void cgroup_kill_sb(struct super_block *sb) {
	struct cgroupfs_root *root = sb->s_fs_info;
	struct cgroup *cgrp = &root->top_cgroup;
	int ret;
	struct cg_cgroup_link *link;
	struct cg_cgroup_link *saved_link;

	BUG_ON(!root);
	// umount时，该层级上的cgroup个数必须为1
	BUG_ON(root->number_of_cgroups != 1);
	// umount时，该层级的root cgroup必须没有子cgroup
	BUG_ON(!list_empty(&cgrp->children));

	mutex_lock(&cgroup_mutex);
	mutex_lock(&cgroup_root_mutex);

	/* Rebind all subsystems back to the default hierarchy */
	// 删除该层级上所有附加的cgroup子系统
	ret = rebind_subsystems(root, 0);
	/* Shouldn't be able to fail ... */
	BUG_ON(ret);

	/*
	 * Release all the links from css_sets to this hierarchy's
	 * root cgroup
	 */
	write_lock(&css_set_lock);

	list_for_each_entry_safe(link, saved_link, &cgrp->css_sets,
				 cgrp_link_list) {
		list_del(&link->cg_link_list);
		list_del(&link->cgrp_link_list);
		kfree(link);
	}
	write_unlock(&css_set_lock);
	// 在roots链表中删除该cgroupfs_root
	if (!list_empty(&root->root_list)) {
		list_del(&root->root_list);
		root_count--;
	}

	mutex_unlock(&cgroup_root_mutex);
	mutex_unlock(&cgroup_mutex);

	simple_xattrs_free(&cgrp->xattrs);

	kill_litter_super(sb);
	cgroup_drop_root(root);
}