[linux-2.6-block.git] / fs / proc / root.c

// SPDX-License-Identifier: GPL-2.0
/*
 *  linux/fs/proc/root.c
 *
 *  Copyright (C) 1991, 1992 Linus Torvalds
 *
 *  proc root directory handling functions
 */

#include <linux/uaccess.h>

#include <linux/errno.h>
#include <linux/time.h>
#include <linux/proc_fs.h>
#include <linux/stat.h>
#include <linux/init.h>
#include <linux/sched.h>
#include <linux/sched/stat.h>
#include <linux/module.h>
#include <linux/bitops.h>
#include <linux/user_namespace.h>
#include <linux/fs_context.h>
#include <linux/mount.h>
#include <linux/pid_namespace.h>
#include <linux/fs_parser.h>
#include <linux/cred.h>
#include <linux/magic.h>
#include <linux/slab.h>

#include "internal.h"

struct proc_fs_context {
	struct pid_namespace	*pid_ns;
	unsigned int		mask;
	int			hidepid;
	int			gid;
};

enum proc_param {
	Opt_gid,
	Opt_hidepid,
};

static const struct fs_parameter_spec proc_param_specs[] = {
	fsparam_u32("gid",	Opt_gid),
	fsparam_u32("hidepid",	Opt_hidepid),
	{}
};

static const struct fs_parameter_description proc_fs_parameters = {
	.name		= "proc",
	.specs		= proc_param_specs,
};

static int proc_parse_param(struct fs_context *fc, struct fs_parameter *param)
{
	struct proc_fs_context *ctx = fc->fs_private;
	struct fs_parse_result result;
	int opt;

	opt = fs_parse(fc, &proc_fs_parameters, param, &result);
	if (opt < 0)
		return opt;

	switch (opt) {
	case Opt_gid:
		ctx->gid = result.uint_32;
		break;

	case Opt_hidepid:
		ctx->hidepid = result.uint_32;
		if (ctx->hidepid < HIDEPID_OFF ||
		    ctx->hidepid > HIDEPID_INVISIBLE)
			return invalf(fc, "proc: hidepid value must be between 0 and 2.\n");
		break;

	default:
		return -EINVAL;
	}

	ctx->mask |= 1 << opt;
	return 0;
}

static void proc_apply_options(struct super_block *s,
			       struct fs_context *fc,
			       struct pid_namespace *pid_ns,
			       struct user_namespace *user_ns)
{
	struct proc_fs_context *ctx = fc->fs_private;

	if (ctx->mask & (1 << Opt_gid))
		pid_ns->pid_gid = make_kgid(user_ns, ctx->gid);
	if (ctx->mask & (1 << Opt_hidepid))
		pid_ns->hide_pid = ctx->hidepid;
}

static int proc_fill_super(struct super_block *s, struct fs_context *fc)
{
	struct pid_namespace *pid_ns = get_pid_ns(s->s_fs_info);
	struct inode *root_inode;
	int ret;

	proc_apply_options(s, fc, pid_ns, current_user_ns());

	/* User space would break if executables or devices appear on proc */
	s->s_iflags |= SB_I_USERNS_VISIBLE | SB_I_NOEXEC | SB_I_NODEV;
	s->s_flags |= SB_NODIRATIME | SB_NOSUID | SB_NOEXEC;
	s->s_blocksize = 1024;
	s->s_blocksize_bits = 10;
	s->s_magic = PROC_SUPER_MAGIC;
	s->s_op = &proc_sops;
	s->s_time_gran = 1;

	/*
	 * procfs isn't actually a stacking filesystem; however, there is
	 * too much magic going on inside it to permit stacking things on
	 * top of it
	 */
	s->s_stack_depth = FILESYSTEM_MAX_STACK_DEPTH;
	
	/* procfs dentries and inodes don't require IO to create */
	s->s_shrink.seeks = 0;

	pde_get(&proc_root);
	root_inode = proc_get_inode(s, &proc_root);
	if (!root_inode) {
		pr_err("proc_fill_super: get root inode failed\n");
		return -ENOMEM;
	}

	s->s_root = d_make_root(root_inode);
	if (!s->s_root) {
		pr_err("proc_fill_super: allocate dentry failed\n");
		return -ENOMEM;
	}

	ret = proc_setup_self(s);
	if (ret) {
		return ret;
	}
	return proc_setup_thread_self(s);
}

static int proc_reconfigure(struct fs_context *fc)
{
	struct super_block *sb = fc->root->d_sb;
	struct pid_namespace *pid = sb->s_fs_info;

	sync_filesystem(sb);

	proc_apply_options(sb, fc, pid, current_user_ns());
	return 0;
}

static int proc_get_tree(struct fs_context *fc)
{
	struct proc_fs_context *ctx = fc->fs_private;

	return get_tree_keyed(fc, proc_fill_super, ctx->pid_ns);
}

static void proc_fs_context_free(struct fs_context *fc)
{
	struct proc_fs_context *ctx = fc->fs_private;

	put_pid_ns(ctx->pid_ns);
	kfree(ctx);
}

static const struct fs_context_operations proc_fs_context_ops = {
	.free		= proc_fs_context_free,
	.parse_param	= proc_parse_param,
	.get_tree	= proc_get_tree,
	.reconfigure	= proc_reconfigure,
};

static int proc_init_fs_context(struct fs_context *fc)
{
	struct proc_fs_context *ctx;

	ctx = kzalloc(sizeof(struct proc_fs_context), GFP_KERNEL);
	if (!ctx)
		return -ENOMEM;

	ctx->pid_ns = get_pid_ns(task_active_pid_ns(current));
	put_user_ns(fc->user_ns);
	fc->user_ns = get_user_ns(ctx->pid_ns->user_ns);
	fc->fs_private = ctx;
	fc->ops = &proc_fs_context_ops;
	return 0;
}

static void proc_kill_sb(struct super_block *sb)
{
	struct pid_namespace *ns;

	ns = (struct pid_namespace *)sb->s_fs_info;
	if (ns->proc_self)
		dput(ns->proc_self);
	if (ns->proc_thread_self)
		dput(ns->proc_thread_self);
	kill_anon_super(sb);
	put_pid_ns(ns);
}

static struct file_system_type proc_fs_type = {
	.name			= "proc",
	.init_fs_context	= proc_init_fs_context,
	.parameters		= &proc_fs_parameters,
	.kill_sb		= proc_kill_sb,
	.fs_flags		= FS_USERNS_MOUNT | FS_DISALLOW_NOTIFY_PERM,
};

void __init proc_root_init(void)
{
	proc_init_kmemcache();
	set_proc_pid_nlink();
	proc_self_init();
	proc_thread_self_init();
	proc_symlink("mounts", NULL, "self/mounts");

	proc_net_init();
	proc_mkdir("fs", NULL);
	proc_mkdir("driver", NULL);
	proc_create_mount_point("fs/nfsd"); /* somewhere for the nfsd filesystem to be mounted */
#if defined(CONFIG_SUN_OPENPROMFS) || defined(CONFIG_SUN_OPENPROMFS_MODULE)
	/* just give it a mountpoint */
	proc_create_mount_point("openprom");
#endif
	proc_tty_init();
	proc_mkdir("bus", NULL);
	proc_sys_init();

	register_filesystem(&proc_fs_type);
}

static int proc_root_getattr(const struct path *path, struct kstat *stat,
			     u32 request_mask, unsigned int query_flags)
{
	generic_fillattr(d_inode(path->dentry), stat);
	stat->nlink = proc_root.nlink + nr_processes();
	return 0;
}

static struct dentry *proc_root_lookup(struct inode * dir, struct dentry * dentry, unsigned int flags)
{
	if (!proc_pid_lookup(dentry, flags))
		return NULL;

	return proc_lookup(dir, dentry, flags);
}

static int proc_root_readdir(struct file *file, struct dir_context *ctx)
{
	if (ctx->pos < FIRST_PROCESS_ENTRY) {
		int error = proc_readdir(file, ctx);
		if (unlikely(error <= 0))
			return error;
		ctx->pos = FIRST_PROCESS_ENTRY;
	}

	return proc_pid_readdir(file, ctx);
}

/*
 * The root /proc directory is special, as it has the
 * <pid> directories. Thus we don't use the generic
 * directory handling functions for that..
 */
static const struct file_operations proc_root_operations = {
	.read		 = generic_read_dir,
	.iterate_shared	 = proc_root_readdir,
	.llseek		= generic_file_llseek,
};

/*
 * proc root can do almost nothing..
 */
static const struct inode_operations proc_root_inode_operations = {
	.lookup		= proc_root_lookup,
	.getattr	= proc_root_getattr,
};

/*
 * This is the root "inode" in the /proc tree..
 */
struct proc_dir_entry proc_root = {
	.low_ino	= PROC_ROOT_INO, 
	.namelen	= 5, 
	.mode		= S_IFDIR | S_IRUGO | S_IXUGO, 
	.nlink		= 2, 
	.refcnt		= REFCOUNT_INIT(1),
	.proc_iops	= &proc_root_inode_operations, 
	.proc_fops	= &proc_root_operations,
	.parent		= &proc_root,
	.subdir		= RB_ROOT,
	.name		= "/proc",
};

int pid_ns_prepare_proc(struct pid_namespace *ns)
{
	struct proc_fs_context *ctx;
	struct fs_context *fc;
	struct vfsmount *mnt;

	fc = fs_context_for_mount(&proc_fs_type, SB_KERNMOUNT);
	if (IS_ERR(fc))
		return PTR_ERR(fc);

	if (fc->user_ns != ns->user_ns) {
		put_user_ns(fc->user_ns);
		fc->user_ns = get_user_ns(ns->user_ns);
	}

	ctx = fc->fs_private;
	if (ctx->pid_ns != ns) {
		put_pid_ns(ctx->pid_ns);
		get_pid_ns(ns);
		ctx->pid_ns = ns;
	}

	mnt = fc_mount(fc);
	put_fs_context(fc);
	if (IS_ERR(mnt))
		return PTR_ERR(mnt);

	ns->proc_mnt = mnt;
	return 0;
}

void pid_ns_release_proc(struct pid_namespace *ns)
{
	kern_unmount(ns->proc_mnt);
}
Commit	Line	Data
b2441318	1	// SPDX-License-Identifier: GPL-2.0
1da177e4 LT	2	/*
	3	* linux/fs/proc/root.c
	4	*
	5	* Copyright (C) 1991, 1992 Linus Torvalds
	6	*
	7	* proc root directory handling functions
	8	*/
	9
7c0f6ba6	10	#include <linux/uaccess.h>
1da177e4 LT	11
	12	#include <linux/errno.h>
	13	#include <linux/time.h>
	14	#include <linux/proc_fs.h>
	15	#include <linux/stat.h>
1da177e4	16	#include <linux/init.h>
914e2637	17	#include <linux/sched.h>
03441a34	18	#include <linux/sched/stat.h>
1da177e4 LT	19	#include <linux/module.h>
1da177e4 LT	20	#include <linux/bitops.h>
87a8ebd6	21	#include <linux/user_namespace.h>
66f592e2	22	#include <linux/fs_context.h>
f6c7a1f3	23	#include <linux/mount.h>
07543f5c	24	#include <linux/pid_namespace.h>
66f592e2	25	#include <linux/fs_parser.h>
5b825c3a	26	#include <linux/cred.h>
60a3c3a5	27	#include <linux/magic.h>
66f592e2	28	#include <linux/slab.h>
1da177e4	29
fee781e6 AB	30	#include "internal.h"
fee781e6 AB	31
66f592e2 DH	32	struct proc_fs_context {
	33	struct pid_namespace *pid_ns;
	34	unsigned int mask;
	35	int hidepid;
	36	int gid;
97412950 VK	37	};
97412950 VK	38
66f592e2 DH	39	enum proc_param {
	40	Opt_gid,
	41	Opt_hidepid,
97412950 VK	42	};
97412950 VK	43
66f592e2 DH	44	static const struct fs_parameter_spec proc_param_specs[] = {
	45	fsparam_u32("gid", Opt_gid),
	46	fsparam_u32("hidepid", Opt_hidepid),
	47	{}
	48	};
	49
	50	static const struct fs_parameter_description proc_fs_parameters = {
	51	.name = "proc",
	52	.specs = proc_param_specs,
	53	};
	54
	55	static int proc_parse_param(struct fs_context fc, struct fs_parameter param)
97412950	56	{
66f592e2 DH	57	struct proc_fs_context *ctx = fc->fs_private;
	58	struct fs_parse_result result;
	59	int opt;
	60
	61	opt = fs_parse(fc, &proc_fs_parameters, param, &result);
	62	if (opt < 0)
	63	return opt;
	64
	65	switch (opt) {
	66	case Opt_gid:
	67	ctx->gid = result.uint_32;
	68	break;
	69
	70	case Opt_hidepid:
	71	ctx->hidepid = result.uint_32;
	72	if (ctx->hidepid < HIDEPID_OFF \|\|
	73	ctx->hidepid > HIDEPID_INVISIBLE)
	74	return invalf(fc, "proc: hidepid value must be between 0 and 2.\n");
	75	break;
	76
	77	default:
	78	return -EINVAL;
97412950 VK	79	}
97412950 VK	80
66f592e2 DH	81	ctx->mask \|= 1 << opt;
66f592e2 DH	82	return 0;
97412950 VK	83	}
97412950 VK	84
66f592e2 DH	85	static void proc_apply_options(struct super_block *s,
	86	struct fs_context *fc,
	87	struct pid_namespace *pid_ns,
	88	struct user_namespace *user_ns)
97412950	89	{
66f592e2 DH	90	struct proc_fs_context *ctx = fc->fs_private;
	91
	92	if (ctx->mask & (1 << Opt_gid))
	93	pid_ns->pid_gid = make_kgid(user_ns, ctx->gid);
	94	if (ctx->mask & (1 << Opt_hidepid))
	95	pid_ns->hide_pid = ctx->hidepid;
97412950 VK	96	}
97412950 VK	97
66f592e2	98	static int proc_fill_super(struct super_block s, struct fs_context fc)
60a3c3a5	99	{
66f592e2	100	struct pid_namespace *pid_ns = get_pid_ns(s->s_fs_info);
60a3c3a5 DH	101	struct inode *root_inode;
	102	int ret;
	103
66f592e2	104	proc_apply_options(s, fc, pid_ns, current_user_ns());
60a3c3a5 DH	105
	106	/* User space would break if executables or devices appear on proc */
	107	s->s_iflags \|= SB_I_USERNS_VISIBLE \| SB_I_NOEXEC \| SB_I_NODEV;
	108	s->s_flags \|= SB_NODIRATIME \| SB_NOSUID \| SB_NOEXEC;
	109	s->s_blocksize = 1024;
	110	s->s_blocksize_bits = 10;
	111	s->s_magic = PROC_SUPER_MAGIC;
	112	s->s_op = &proc_sops;
	113	s->s_time_gran = 1;
	114
	115	/*
	116	* procfs isn't actually a stacking filesystem; however, there is
	117	* too much magic going on inside it to permit stacking things on
	118	* top of it
	119	*/
	120	s->s_stack_depth = FILESYSTEM_MAX_STACK_DEPTH;
	121
	122	/* procfs dentries and inodes don't require IO to create */
	123	s->s_shrink.seeks = 0;
	124
	125	pde_get(&proc_root);
	126	root_inode = proc_get_inode(s, &proc_root);
	127	if (!root_inode) {
	128	pr_err("proc_fill_super: get root inode failed\n");
	129	return -ENOMEM;
	130	}
	131
	132	s->s_root = d_make_root(root_inode);
	133	if (!s->s_root) {
	134	pr_err("proc_fill_super: allocate dentry failed\n");
	135	return -ENOMEM;
	136	}
	137
	138	ret = proc_setup_self(s);
	139	if (ret) {
	140	return ret;
	141	}
	142	return proc_setup_thread_self(s);
	143	}
	144
66f592e2	145	static int proc_reconfigure(struct fs_context *fc)
97412950	146	{
66f592e2	147	struct super_block *sb = fc->root->d_sb;
97412950	148	struct pid_namespace *pid = sb->s_fs_info;
02b9984d TT	149
02b9984d TT	150	sync_filesystem(sb);
66f592e2 DH	151
	152	proc_apply_options(sb, fc, pid, current_user_ns());
	153	return 0;
97412950 VK	154	}
97412950 VK	155
66f592e2	156	static int proc_get_tree(struct fs_context *fc)
1da177e4	157	{
66f592e2	158	struct proc_fs_context *ctx = fc->fs_private;
07543f5c	159
533770cc	160	return get_tree_keyed(fc, proc_fill_super, ctx->pid_ns);
66f592e2	161	}
07543f5c	162
66f592e2 DH	163	static void proc_fs_context_free(struct fs_context *fc)
	164	{
	165	struct proc_fs_context *ctx = fc->fs_private;
	166
46cf047a	167	put_pid_ns(ctx->pid_ns);
66f592e2 DH	168	kfree(ctx);
	169	}
	170
	171	static const struct fs_context_operations proc_fs_context_ops = {
	172	.free = proc_fs_context_free,
	173	.parse_param = proc_parse_param,
	174	.get_tree = proc_get_tree,
	175	.reconfigure = proc_reconfigure,
	176	};
	177
	178	static int proc_init_fs_context(struct fs_context *fc)
	179	{
	180	struct proc_fs_context *ctx;
	181
	182	ctx = kzalloc(sizeof(struct proc_fs_context), GFP_KERNEL);
	183	if (!ctx)
	184	return -ENOMEM;
07543f5c	185
66f592e2	186	ctx->pid_ns = get_pid_ns(task_active_pid_ns(current));
46cf047a AV	187	put_user_ns(fc->user_ns);
46cf047a AV	188	fc->user_ns = get_user_ns(ctx->pid_ns->user_ns);
66f592e2 DH	189	fc->fs_private = ctx;
	190	fc->ops = &proc_fs_context_ops;
	191	return 0;
07543f5c PE	192	}
	193
	194	static void proc_kill_sb(struct super_block *sb)
	195	{
	196	struct pid_namespace *ns;
	197
	198	ns = (struct pid_namespace *)sb->s_fs_info;
021ada7d AV	199	if (ns->proc_self)
021ada7d AV	200	dput(ns->proc_self);
0097875b EB	201	if (ns->proc_thread_self)
0097875b EB	202	dput(ns->proc_thread_self);
07543f5c PE	203	kill_anon_super(sb);
07543f5c PE	204	put_pid_ns(ns);
1da177e4 LT	205	}
1da177e4 LT	206
c2319540	207	static struct file_system_type proc_fs_type = {
66f592e2 DH	208	.name = "proc",
	209	.init_fs_context = proc_init_fs_context,
	210	.parameters = &proc_fs_parameters,
	211	.kill_sb = proc_kill_sb,
0b3b094a	212	.fs_flags = FS_USERNS_MOUNT \| FS_DISALLOW_NOTIFY_PERM,
1da177e4 LT	213	};
1da177e4 LT	214
1da177e4 LT	215	void __init proc_root_init(void)
1da177e4 LT	216	{
195b8cf0	217	proc_init_kmemcache();
1270dd8d	218	set_proc_pid_nlink();
e656d8a6	219	proc_self_init();
0097875b	220	proc_thread_self_init();
155134fe	221	proc_symlink("mounts", NULL, "self/mounts");
457c4cbc EB	222
457c4cbc EB	223	proc_net_init();
36a5aeb8	224	proc_mkdir("fs", NULL);
928b4d8c	225	proc_mkdir("driver", NULL);
eb6d38d5	226	proc_create_mount_point("fs/nfsd"); /* somewhere for the nfsd filesystem to be mounted */
1da177e4 LT	227	#if defined(CONFIG_SUN_OPENPROMFS) \|\| defined(CONFIG_SUN_OPENPROMFS_MODULE)
1da177e4 LT	228	/* just give it a mountpoint */
eb6d38d5	229	proc_create_mount_point("openprom");
1da177e4 LT	230	#endif
1da177e4 LT	231	proc_tty_init();
9c37066d	232	proc_mkdir("bus", NULL);
77b14db5	233	proc_sys_init();
1539d584 AD	234
1539d584 AD	235	register_filesystem(&proc_fs_type);
1da177e4 LT	236	}
1da177e4 LT	237
a528d35e DH	238	static int proc_root_getattr(const struct path path, struct kstat stat,
a528d35e DH	239	u32 request_mask, unsigned int query_flags)
1da177e4	240	{
a528d35e	241	generic_fillattr(d_inode(path->dentry), stat);
76b6159b AV	242	stat->nlink = proc_root.nlink + nr_processes();
	243	return 0;
	244	}
1da177e4	245
00cd8dd3	246	static struct dentry proc_root_lookup(struct inode dir, struct dentry * dentry, unsigned int flags)
76b6159b	247	{
867aaccf	248	if (!proc_pid_lookup(dentry, flags))
1da177e4	249	return NULL;
66f592e2	250
335eb531	251	return proc_lookup(dir, dentry, flags);
1da177e4 LT	252	}
1da177e4 LT	253
f0c3b509	254	static int proc_root_readdir(struct file file, struct dir_context ctx)
1da177e4	255	{
f0c3b509	256	if (ctx->pos < FIRST_PROCESS_ENTRY) {
94fc5d9d RG	257	int error = proc_readdir(file, ctx);
	258	if (unlikely(error <= 0))
	259	return error;
f0c3b509	260	ctx->pos = FIRST_PROCESS_ENTRY;
1da177e4	261	}
1da177e4	262
f0c3b509	263	return proc_pid_readdir(file, ctx);
1da177e4 LT	264	}
	265
	266	/*
	267	* The root /proc directory is special, as it has the
	268	* <pid> directories. Thus we don't use the generic
	269	* directory handling functions for that..
	270	*/
00977a59	271	static const struct file_operations proc_root_operations = {
1da177e4	272	.read = generic_read_dir,
f50752ea AV	273	.iterate_shared = proc_root_readdir,
f50752ea AV	274	.llseek = generic_file_llseek,
1da177e4 LT	275	};
	276
	277	/*
	278	* proc root can do almost nothing..
	279	*/
c5ef1c42	280	static const struct inode_operations proc_root_inode_operations = {
1da177e4	281	.lookup = proc_root_lookup,
76b6159b	282	.getattr = proc_root_getattr,
1da177e4 LT	283	};
	284
	285	/*
	286	* This is the root "inode" in the /proc tree..
	287	*/
	288	struct proc_dir_entry proc_root = {
	289	.low_ino = PROC_ROOT_INO,
	290	.namelen = 5,
1da177e4 LT	291	.mode = S_IFDIR \| S_IRUGO \| S_IXUGO,
1da177e4 LT	292	.nlink = 2,
9cdd83e3	293	.refcnt = REFCOUNT_INIT(1),
1da177e4 LT	294	.proc_iops = &proc_root_inode_operations,
	295	.proc_fops = &proc_root_operations,
	296	.parent = &proc_root,
4f113437	297	.subdir = RB_ROOT,
24074a35	298	.name = "/proc",
1da177e4 LT	299	};
1da177e4 LT	300
6f4e6433 PE	301	int pid_ns_prepare_proc(struct pid_namespace *ns)
6f4e6433 PE	302	{
66f592e2 DH	303	struct proc_fs_context *ctx;
66f592e2 DH	304	struct fs_context *fc;
6f4e6433 PE	305	struct vfsmount *mnt;
6f4e6433 PE	306
66f592e2 DH	307	fc = fs_context_for_mount(&proc_fs_type, SB_KERNMOUNT);
	308	if (IS_ERR(fc))
	309	return PTR_ERR(fc);
	310
	311	if (fc->user_ns != ns->user_ns) {
	312	put_user_ns(fc->user_ns);
	313	fc->user_ns = get_user_ns(ns->user_ns);
	314	}
	315
	316	ctx = fc->fs_private;
	317	if (ctx->pid_ns != ns) {
	318	put_pid_ns(ctx->pid_ns);
	319	get_pid_ns(ns);
	320	ctx->pid_ns = ns;
	321	}
	322
	323	mnt = fc_mount(fc);
	324	put_fs_context(fc);
6f4e6433 PE	325	if (IS_ERR(mnt))
	326	return PTR_ERR(mnt);
	327
579441a3	328	ns->proc_mnt = mnt;
6f4e6433 PE	329	return 0;
	330	}
	331
	332	void pid_ns_release_proc(struct pid_namespace *ns)
	333	{
905ad269	334	kern_unmount(ns->proc_mnt);
6f4e6433	335	}