[linux-block.git] / fs / proc / inode.c

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

#include <linux/cache.h>
#include <linux/time.h>
#include <linux/proc_fs.h>
#include <linux/kernel.h>
#include <linux/pid_namespace.h>
#include <linux/mm.h>
#include <linux/string.h>
#include <linux/stat.h>
#include <linux/completion.h>
#include <linux/poll.h>
#include <linux/printk.h>
#include <linux/file.h>
#include <linux/limits.h>
#include <linux/init.h>
#include <linux/module.h>
#include <linux/sysctl.h>
#include <linux/seq_file.h>
#include <linux/slab.h>
#include <linux/mount.h>

#include <linux/uaccess.h>

#include "internal.h"

static void proc_evict_inode(struct inode *inode)
{
	struct proc_dir_entry *de;
	struct ctl_table_header *head;
	struct proc_inode *ei = PROC_I(inode);

	truncate_inode_pages_final(&inode->i_data);
	clear_inode(inode);

	/* Stop tracking associated processes */
	if (ei->pid) {
		proc_pid_evict_inode(ei);
		ei->pid = NULL;
	}

	/* Let go of any associated proc directory entry */
	de = ei->pde;
	if (de) {
		pde_put(de);
		ei->pde = NULL;
	}

	head = ei->sysctl;
	if (head) {
		RCU_INIT_POINTER(ei->sysctl, NULL);
		proc_sys_evict_inode(inode, head);
	}
}

static struct kmem_cache *proc_inode_cachep __ro_after_init;
static struct kmem_cache *pde_opener_cache __ro_after_init;

static struct inode *proc_alloc_inode(struct super_block *sb)
{
	struct proc_inode *ei;

	ei = kmem_cache_alloc(proc_inode_cachep, GFP_KERNEL);
	if (!ei)
		return NULL;
	ei->pid = NULL;
	ei->fd = 0;
	ei->op.proc_get_link = NULL;
	ei->pde = NULL;
	ei->sysctl = NULL;
	ei->sysctl_entry = NULL;
	INIT_HLIST_NODE(&ei->sibling_inodes);
	ei->ns_ops = NULL;
	return &ei->vfs_inode;
}

static void proc_free_inode(struct inode *inode)
{
	kmem_cache_free(proc_inode_cachep, PROC_I(inode));
}

static void init_once(void *foo)
{
	struct proc_inode *ei = (struct proc_inode *) foo;

	inode_init_once(&ei->vfs_inode);
}

void __init proc_init_kmemcache(void)
{
	proc_inode_cachep = kmem_cache_create("proc_inode_cache",
					     sizeof(struct proc_inode),
					     0, (SLAB_RECLAIM_ACCOUNT|
						SLAB_MEM_SPREAD|SLAB_ACCOUNT|
						SLAB_PANIC),
					     init_once);
	pde_opener_cache =
		kmem_cache_create("pde_opener", sizeof(struct pde_opener), 0,
				  SLAB_ACCOUNT|SLAB_PANIC, NULL);
	proc_dir_entry_cache = kmem_cache_create_usercopy(
		"proc_dir_entry", SIZEOF_PDE, 0, SLAB_PANIC,
		offsetof(struct proc_dir_entry, inline_name),
		SIZEOF_PDE_INLINE_NAME, NULL);
	BUILD_BUG_ON(sizeof(struct proc_dir_entry) >= SIZEOF_PDE);
}

void proc_invalidate_siblings_dcache(struct hlist_head *inodes, spinlock_t *lock)
{
	struct inode *inode;
	struct proc_inode *ei;
	struct hlist_node *node;
	struct super_block *old_sb = NULL;

	rcu_read_lock();
	for (;;) {
		struct super_block *sb;
		node = hlist_first_rcu(inodes);
		if (!node)
			break;
		ei = hlist_entry(node, struct proc_inode, sibling_inodes);
		spin_lock(lock);
		hlist_del_init_rcu(&ei->sibling_inodes);
		spin_unlock(lock);

		inode = &ei->vfs_inode;
		sb = inode->i_sb;
		if ((sb != old_sb) && !atomic_inc_not_zero(&sb->s_active))
			continue;
		inode = igrab(inode);
		rcu_read_unlock();
		if (sb != old_sb) {
			if (old_sb)
				deactivate_super(old_sb);
			old_sb = sb;
		}
		if (unlikely(!inode)) {
			rcu_read_lock();
			continue;
		}

		if (S_ISDIR(inode->i_mode)) {
			struct dentry *dir = d_find_any_alias(inode);
			if (dir) {
				d_invalidate(dir);
				dput(dir);
			}
		} else {
			struct dentry *dentry;
			while ((dentry = d_find_alias(inode))) {
				d_invalidate(dentry);
				dput(dentry);
			}
		}
		iput(inode);

		rcu_read_lock();
	}
	rcu_read_unlock();
	if (old_sb)
		deactivate_super(old_sb);
}

static int proc_show_options(struct seq_file *seq, struct dentry *root)
{
	struct super_block *sb = root->d_sb;
	struct pid_namespace *pid = sb->s_fs_info;

	if (!gid_eq(pid->pid_gid, GLOBAL_ROOT_GID))
		seq_printf(seq, ",gid=%u", from_kgid_munged(&init_user_ns, pid->pid_gid));
	if (pid->hide_pid != HIDEPID_OFF)
		seq_printf(seq, ",hidepid=%u", pid->hide_pid);

	return 0;
}

const struct super_operations proc_sops = {
	.alloc_inode	= proc_alloc_inode,
	.free_inode	= proc_free_inode,
	.drop_inode	= generic_delete_inode,
	.evict_inode	= proc_evict_inode,
	.statfs		= simple_statfs,
	.show_options	= proc_show_options,
};

enum {BIAS = -1U<<31};

static inline int use_pde(struct proc_dir_entry *pde)
{
	return likely(atomic_inc_unless_negative(&pde->in_use));
}

static void unuse_pde(struct proc_dir_entry *pde)
{
	if (unlikely(atomic_dec_return(&pde->in_use) == BIAS))
		complete(pde->pde_unload_completion);
}

/* pde is locked on entry, unlocked on exit */
static void close_pdeo(struct proc_dir_entry *pde, struct pde_opener *pdeo)
{
	/*
	 * close() (proc_reg_release()) can't delete an entry and proceed:
	 * ->release hook needs to be available at the right moment.
	 *
	 * rmmod (remove_proc_entry() et al) can't delete an entry and proceed:
	 * "struct file" needs to be available at the right moment.
	 *
	 * Therefore, first process to enter this function does ->release() and
	 * signals its completion to the other process which does nothing.
	 */
	if (pdeo->closing) {
		/* somebody else is doing that, just wait */
		DECLARE_COMPLETION_ONSTACK(c);
		pdeo->c = &c;
		spin_unlock(&pde->pde_unload_lock);
		wait_for_completion(&c);
	} else {
		struct file *file;
		struct completion *c;

		pdeo->closing = true;
		spin_unlock(&pde->pde_unload_lock);
		file = pdeo->file;
		pde->proc_ops->proc_release(file_inode(file), file);
		spin_lock(&pde->pde_unload_lock);
		/* After ->release. */
		list_del(&pdeo->lh);
		c = pdeo->c;
		spin_unlock(&pde->pde_unload_lock);
		if (unlikely(c))
			complete(c);
		kmem_cache_free(pde_opener_cache, pdeo);
	}
}

void proc_entry_rundown(struct proc_dir_entry *de)
{
	DECLARE_COMPLETION_ONSTACK(c);
	/* Wait until all existing callers into module are done. */
	de->pde_unload_completion = &c;
	if (atomic_add_return(BIAS, &de->in_use) != BIAS)
		wait_for_completion(&c);

	/* ->pde_openers list can't grow from now on. */

	spin_lock(&de->pde_unload_lock);
	while (!list_empty(&de->pde_openers)) {
		struct pde_opener *pdeo;
		pdeo = list_first_entry(&de->pde_openers, struct pde_opener, lh);
		close_pdeo(de, pdeo);
		spin_lock(&de->pde_unload_lock);
	}
	spin_unlock(&de->pde_unload_lock);
}

static loff_t proc_reg_llseek(struct file *file, loff_t offset, int whence)
{
	struct proc_dir_entry *pde = PDE(file_inode(file));
	loff_t rv = -EINVAL;
	if (use_pde(pde)) {
		typeof_member(struct proc_ops, proc_lseek) lseek;

		lseek = pde->proc_ops->proc_lseek;
		if (!lseek)
			lseek = default_llseek;
		rv = lseek(file, offset, whence);
		unuse_pde(pde);
	}
	return rv;
}

static ssize_t proc_reg_read(struct file *file, char __user *buf, size_t count, loff_t *ppos)
{
	struct proc_dir_entry *pde = PDE(file_inode(file));
	ssize_t rv = -EIO;
	if (use_pde(pde)) {
		typeof_member(struct proc_ops, proc_read) read;

		read = pde->proc_ops->proc_read;
		if (read)
			rv = read(file, buf, count, ppos);
		unuse_pde(pde);
	}
	return rv;
}

static ssize_t proc_reg_write(struct file *file, const char __user *buf, size_t count, loff_t *ppos)
{
	struct proc_dir_entry *pde = PDE(file_inode(file));
	ssize_t rv = -EIO;
	if (use_pde(pde)) {
		typeof_member(struct proc_ops, proc_write) write;

		write = pde->proc_ops->proc_write;
		if (write)
			rv = write(file, buf, count, ppos);
		unuse_pde(pde);
	}
	return rv;
}

static __poll_t proc_reg_poll(struct file *file, struct poll_table_struct *pts)
{
	struct proc_dir_entry *pde = PDE(file_inode(file));
	__poll_t rv = DEFAULT_POLLMASK;
	if (use_pde(pde)) {
		typeof_member(struct proc_ops, proc_poll) poll;

		poll = pde->proc_ops->proc_poll;
		if (poll)
			rv = poll(file, pts);
		unuse_pde(pde);
	}
	return rv;
}

static long proc_reg_unlocked_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
{
	struct proc_dir_entry *pde = PDE(file_inode(file));
	long rv = -ENOTTY;
	if (use_pde(pde)) {
		typeof_member(struct proc_ops, proc_ioctl) ioctl;

		ioctl = pde->proc_ops->proc_ioctl;
		if (ioctl)
			rv = ioctl(file, cmd, arg);
		unuse_pde(pde);
	}
	return rv;
}

#ifdef CONFIG_COMPAT
static long proc_reg_compat_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
{
	struct proc_dir_entry *pde = PDE(file_inode(file));
	long rv = -ENOTTY;
	if (use_pde(pde)) {
		typeof_member(struct proc_ops, proc_compat_ioctl) compat_ioctl;

		compat_ioctl = pde->proc_ops->proc_compat_ioctl;
		if (compat_ioctl)
			rv = compat_ioctl(file, cmd, arg);
		unuse_pde(pde);
	}
	return rv;
}
#endif

static int proc_reg_mmap(struct file *file, struct vm_area_struct *vma)
{
	struct proc_dir_entry *pde = PDE(file_inode(file));
	int rv = -EIO;
	if (use_pde(pde)) {
		typeof_member(struct proc_ops, proc_mmap) mmap;

		mmap = pde->proc_ops->proc_mmap;
		if (mmap)
			rv = mmap(file, vma);
		unuse_pde(pde);
	}
	return rv;
}

static unsigned long
proc_reg_get_unmapped_area(struct file *file, unsigned long orig_addr,
			   unsigned long len, unsigned long pgoff,
			   unsigned long flags)
{
	struct proc_dir_entry *pde = PDE(file_inode(file));
	unsigned long rv = -EIO;

	if (use_pde(pde)) {
		typeof_member(struct proc_ops, proc_get_unmapped_area) get_area;

		get_area = pde->proc_ops->proc_get_unmapped_area;
#ifdef CONFIG_MMU
		if (!get_area)
			get_area = current->mm->get_unmapped_area;
#endif

		if (get_area)
			rv = get_area(file, orig_addr, len, pgoff, flags);
		else
			rv = orig_addr;
		unuse_pde(pde);
	}
	return rv;
}

static int proc_reg_open(struct inode *inode, struct file *file)
{
	struct proc_dir_entry *pde = PDE(inode);
	int rv = 0;
	typeof_member(struct proc_ops, proc_open) open;
	typeof_member(struct proc_ops, proc_release) release;
	struct pde_opener *pdeo;

	/*
	 * Ensure that
	 * 1) PDE's ->release hook will be called no matter what
	 *    either normally by close()/->release, or forcefully by
	 *    rmmod/remove_proc_entry.
	 *
	 * 2) rmmod isn't blocked by opening file in /proc and sitting on
	 *    the descriptor (including "rmmod foo </proc/foo" scenario).
	 *
	 * Save every "struct file" with custom ->release hook.
	 */
	if (!use_pde(pde))
		return -ENOENT;

	release = pde->proc_ops->proc_release;
	if (release) {
		pdeo = kmem_cache_alloc(pde_opener_cache, GFP_KERNEL);
		if (!pdeo) {
			rv = -ENOMEM;
			goto out_unuse;
		}
	}

	open = pde->proc_ops->proc_open;
	if (open)
		rv = open(inode, file);

	if (release) {
		if (rv == 0) {
			/* To know what to release. */
			pdeo->file = file;
			pdeo->closing = false;
			pdeo->c = NULL;
			spin_lock(&pde->pde_unload_lock);
			list_add(&pdeo->lh, &pde->pde_openers);
			spin_unlock(&pde->pde_unload_lock);
		} else
			kmem_cache_free(pde_opener_cache, pdeo);
	}

out_unuse:
	unuse_pde(pde);
	return rv;
}

static int proc_reg_release(struct inode *inode, struct file *file)
{
	struct proc_dir_entry *pde = PDE(inode);
	struct pde_opener *pdeo;
	spin_lock(&pde->pde_unload_lock);
	list_for_each_entry(pdeo, &pde->pde_openers, lh) {
		if (pdeo->file == file) {
			close_pdeo(pde, pdeo);
			return 0;
		}
	}
	spin_unlock(&pde->pde_unload_lock);
	return 0;
}

static const struct file_operations proc_reg_file_ops = {
	.llseek		= proc_reg_llseek,
	.read		= proc_reg_read,
	.write		= proc_reg_write,
	.poll		= proc_reg_poll,
	.unlocked_ioctl	= proc_reg_unlocked_ioctl,
#ifdef CONFIG_COMPAT
	.compat_ioctl	= proc_reg_compat_ioctl,
#endif
	.mmap		= proc_reg_mmap,
	.get_unmapped_area = proc_reg_get_unmapped_area,
	.open		= proc_reg_open,
	.release	= proc_reg_release,
};

#ifdef CONFIG_COMPAT
static const struct file_operations proc_reg_file_ops_no_compat = {
	.llseek		= proc_reg_llseek,
	.read		= proc_reg_read,
	.write		= proc_reg_write,
	.poll		= proc_reg_poll,
	.unlocked_ioctl	= proc_reg_unlocked_ioctl,
	.mmap		= proc_reg_mmap,
	.get_unmapped_area = proc_reg_get_unmapped_area,
	.open		= proc_reg_open,
	.release	= proc_reg_release,
};
#endif

static void proc_put_link(void *p)
{
	unuse_pde(p);
}

static const char *proc_get_link(struct dentry *dentry,
				 struct inode *inode,
				 struct delayed_call *done)
{
	struct proc_dir_entry *pde = PDE(inode);
	if (!use_pde(pde))
		return ERR_PTR(-EINVAL);
	set_delayed_call(done, proc_put_link, pde);
	return pde->data;
}

const struct inode_operations proc_link_inode_operations = {
	.get_link	= proc_get_link,
};

struct inode *proc_get_inode(struct super_block *sb, struct proc_dir_entry *de)
{
	struct inode *inode = new_inode_pseudo(sb);

	if (inode) {
		inode->i_ino = de->low_ino;
		inode->i_mtime = inode->i_atime = inode->i_ctime = current_time(inode);
		PROC_I(inode)->pde = de;

		if (is_empty_pde(de)) {
			make_empty_dir_inode(inode);
			return inode;
		}
		if (de->mode) {
			inode->i_mode = de->mode;
			inode->i_uid = de->uid;
			inode->i_gid = de->gid;
		}
		if (de->size)
			inode->i_size = de->size;
		if (de->nlink)
			set_nlink(inode, de->nlink);

		if (S_ISREG(inode->i_mode)) {
			inode->i_op = de->proc_iops;
			inode->i_fop = &proc_reg_file_ops;
#ifdef CONFIG_COMPAT
			if (!de->proc_ops->proc_compat_ioctl) {
				inode->i_fop = &proc_reg_file_ops_no_compat;
			}
#endif
		} else if (S_ISDIR(inode->i_mode)) {
			inode->i_op = de->proc_iops;
			inode->i_fop = de->proc_dir_ops;
		} else if (S_ISLNK(inode->i_mode)) {
			inode->i_op = de->proc_iops;
			inode->i_fop = NULL;
		} else
			BUG();
	} else
	       pde_put(de);
	return inode;
}
Commit	Line	Data
b2441318	1	// SPDX-License-Identifier: GPL-2.0
1da177e4 LT	2	/*
	3	* linux/fs/proc/inode.c
	4	*
	5	* Copyright (C) 1991, 1992 Linus Torvalds
	6	*/
	7
efb1a57d	8	#include <linux/cache.h>
1da177e4 LT	9	#include <linux/time.h>
	10	#include <linux/proc_fs.h>
	11	#include <linux/kernel.h>
97412950	12	#include <linux/pid_namespace.h>
1da177e4 LT	13	#include <linux/mm.h>
	14	#include <linux/string.h>
	15	#include <linux/stat.h>
786d7e16	16	#include <linux/completion.h>
dd23aae4	17	#include <linux/poll.h>
87ebdc00	18	#include <linux/printk.h>
1da177e4 LT	19	#include <linux/file.h>
	20	#include <linux/limits.h>
	21	#include <linux/init.h>
	22	#include <linux/module.h>
9043476f	23	#include <linux/sysctl.h>
97412950	24	#include <linux/seq_file.h>
5a0e3ad6	25	#include <linux/slab.h>
97412950	26	#include <linux/mount.h>
1da177e4	27
7c0f6ba6	28	#include <linux/uaccess.h>
1da177e4	29
fee781e6	30	#include "internal.h"
1da177e4	31
8267952b	32	static void proc_evict_inode(struct inode *inode)
1da177e4 LT	33	{
1da177e4 LT	34	struct proc_dir_entry *de;
dfef6dcd	35	struct ctl_table_header *head;
71448011	36	struct proc_inode *ei = PROC_I(inode);
1da177e4	37
91b0abe3	38	truncate_inode_pages_final(&inode->i_data);
dbd5768f	39	clear_inode(inode);
fef26658	40
99f89551	41	/* Stop tracking associated processes */
71448011	42	if (ei->pid) {
7bc3e6e5	43	proc_pid_evict_inode(ei);
71448011 EB	44	ei->pid = NULL;
71448011 EB	45	}
1da177e4 LT	46
1da177e4 LT	47	/* Let go of any associated proc directory entry */
71448011 EB	48	de = ei->pde;
71448011 EB	49	if (de) {
135d5655	50	pde_put(de);
71448011 EB	51	ei->pde = NULL;
71448011 EB	52	}
d6cffbbe	53
71448011	54	head = ei->sysctl;
dfef6dcd	55	if (head) {
71448011	56	RCU_INIT_POINTER(ei->sysctl, NULL);
d6cffbbe	57	proc_sys_evict_inode(inode, head);
dfef6dcd	58	}
1da177e4 LT	59	}
1da177e4 LT	60
efb1a57d	61	static struct kmem_cache *proc_inode_cachep __ro_after_init;
195b8cf0	62	static struct kmem_cache *pde_opener_cache __ro_after_init;
1da177e4 LT	63
	64	static struct inode proc_alloc_inode(struct super_block sb)
	65	{
	66	struct proc_inode *ei;
1da177e4	67
f245e1c1	68	ei = kmem_cache_alloc(proc_inode_cachep, GFP_KERNEL);
1da177e4 LT	69	if (!ei)
1da177e4 LT	70	return NULL;
13b41b09	71	ei->pid = NULL;
aed7a6c4	72	ei->fd = 0;
1da177e4 LT	73	ei->op.proc_get_link = NULL;
1da177e4 LT	74	ei->pde = NULL;
9043476f AV	75	ei->sysctl = NULL;
9043476f AV	76	ei->sysctl_entry = NULL;
0afa5ca8	77	INIT_HLIST_NODE(&ei->sibling_inodes);
3d3d35b1	78	ei->ns_ops = NULL;
230f72e9	79	return &ei->vfs_inode;
1da177e4 LT	80	}
1da177e4 LT	81
4aa6b55c	82	static void proc_free_inode(struct inode *inode)
1da177e4 LT	83	{
	84	kmem_cache_free(proc_inode_cachep, PROC_I(inode));
	85	}
	86
51cc5068	87	static void init_once(void *foo)
1da177e4 LT	88	{
	89	struct proc_inode ei = (struct proc_inode ) foo;
	90
a35afb83	91	inode_init_once(&ei->vfs_inode);
1da177e4	92	}
20c2df83	93
195b8cf0	94	void __init proc_init_kmemcache(void)
1da177e4 LT	95	{
	96	proc_inode_cachep = kmem_cache_create("proc_inode_cache",
	97	sizeof(struct proc_inode),
fffb60f9	98	0, (SLAB_RECLAIM_ACCOUNT\|
5d097056 VD	99	SLAB_MEM_SPREAD\|SLAB_ACCOUNT\|
5d097056 VD	100	SLAB_PANIC),
20c2df83	101	init_once);
195b8cf0 AD	102	pde_opener_cache =
195b8cf0 AD	103	kmem_cache_create("pde_opener", sizeof(struct pde_opener), 0,
2acddbe8	104	SLAB_ACCOUNT\|SLAB_PANIC, NULL);
b4884f23	105	proc_dir_entry_cache = kmem_cache_create_usercopy(
2d6e4e82 AD	106	"proc_dir_entry", SIZEOF_PDE, 0, SLAB_PANIC,
	107	offsetof(struct proc_dir_entry, inline_name),
	108	SIZEOF_PDE_INLINE_NAME, NULL);
	109	BUILD_BUG_ON(sizeof(struct proc_dir_entry) >= SIZEOF_PDE);
1da177e4 LT	110	}
1da177e4 LT	111
f90f3caf	112	void proc_invalidate_siblings_dcache(struct hlist_head inodes, spinlock_t lock)
26dbc60f EB	113	{
	114	struct inode *inode;
	115	struct proc_inode *ei;
	116	struct hlist_node *node;
080f6276	117	struct super_block *old_sb = NULL;
26dbc60f EB	118
	119	rcu_read_lock();
	120	for (;;) {
080f6276	121	struct super_block *sb;
26dbc60f EB	122	node = hlist_first_rcu(inodes);
	123	if (!node)
	124	break;
	125	ei = hlist_entry(node, struct proc_inode, sibling_inodes);
	126	spin_lock(lock);
	127	hlist_del_init_rcu(&ei->sibling_inodes);
	128	spin_unlock(lock);
	129
	130	inode = &ei->vfs_inode;
	131	sb = inode->i_sb;
080f6276	132	if ((sb != old_sb) && !atomic_inc_not_zero(&sb->s_active))
26dbc60f EB	133	continue;
	134	inode = igrab(inode);
	135	rcu_read_unlock();
080f6276 EB	136	if (sb != old_sb) {
	137	if (old_sb)
	138	deactivate_super(old_sb);
	139	old_sb = sb;
	140	}
26dbc60f	141	if (unlikely(!inode)) {
26dbc60f EB	142	rcu_read_lock();
	143	continue;
	144	}
	145
f90f3caf EB	146	if (S_ISDIR(inode->i_mode)) {
	147	struct dentry *dir = d_find_any_alias(inode);
	148	if (dir) {
	149	d_invalidate(dir);
	150	dput(dir);
	151	}
	152	} else {
	153	struct dentry *dentry;
	154	while ((dentry = d_find_alias(inode))) {
	155	d_invalidate(dentry);
	156	dput(dentry);
	157	}
	158	}
26dbc60f	159	iput(inode);
26dbc60f EB	160
	161	rcu_read_lock();
	162	}
	163	rcu_read_unlock();
080f6276 EB	164	if (old_sb)
080f6276 EB	165	deactivate_super(old_sb);
26dbc60f EB	166	}
26dbc60f EB	167
97412950 VK	168	static int proc_show_options(struct seq_file seq, struct dentry root)
97412950 VK	169	{
0499680a VK	170	struct super_block *sb = root->d_sb;
	171	struct pid_namespace *pid = sb->s_fs_info;
	172
dcb0f222 EB	173	if (!gid_eq(pid->pid_gid, GLOBAL_ROOT_GID))
dcb0f222 EB	174	seq_printf(seq, ",gid=%u", from_kgid_munged(&init_user_ns, pid->pid_gid));
796f571b	175	if (pid->hide_pid != HIDEPID_OFF)
0499680a VK	176	seq_printf(seq, ",hidepid=%u", pid->hide_pid);
0499680a VK	177
97412950 VK	178	return 0;
	179	}
	180
60a3c3a5	181	const struct super_operations proc_sops = {
1da177e4	182	.alloc_inode = proc_alloc_inode,
4aa6b55c	183	.free_inode = proc_free_inode,
1da177e4	184	.drop_inode = generic_delete_inode,
8267952b	185	.evict_inode = proc_evict_inode,
1da177e4	186	.statfs = simple_statfs,
97412950	187	.show_options = proc_show_options,
1da177e4 LT	188	};
1da177e4 LT	189
866ad9a7 AV	190	enum {BIAS = -1U<<31};
	191
	192	static inline int use_pde(struct proc_dir_entry *pde)
	193	{
15b158b4	194	return likely(atomic_inc_unless_negative(&pde->in_use));
881adb85 AD	195	}
881adb85 AD	196
866ad9a7	197	static void unuse_pde(struct proc_dir_entry *pde)
881adb85	198	{
15b158b4	199	if (unlikely(atomic_dec_return(&pde->in_use) == BIAS))
05c0ae21	200	complete(pde->pde_unload_completion);
786d7e16 AD	201	}
786d7e16 AD	202
2f897424	203	/* pde is locked on entry, unlocked on exit */
ca469f35 AV	204	static void close_pdeo(struct proc_dir_entry pde, struct pde_opener pdeo)
ca469f35 AV	205	{
492b2da6 AD	206	/*
	207	* close() (proc_reg_release()) can't delete an entry and proceed:
	208	* ->release hook needs to be available at the right moment.
	209	*
	210	* rmmod (remove_proc_entry() et al) can't delete an entry and proceed:
	211	* "struct file" needs to be available at the right moment.
	212	*
	213	* Therefore, first process to enter this function does ->release() and
	214	* signals its completion to the other process which does nothing.
	215	*/
05c0ae21	216	if (pdeo->closing) {
ca469f35	217	/* somebody else is doing that, just wait */
05c0ae21 AV	218	DECLARE_COMPLETION_ONSTACK(c);
05c0ae21 AV	219	pdeo->c = &c;
ca469f35	220	spin_unlock(&pde->pde_unload_lock);
05c0ae21	221	wait_for_completion(&c);
ca469f35 AV	222	} else {
ca469f35 AV	223	struct file *file;
2f897424 AD	224	struct completion *c;
2f897424 AD	225
f5887c71	226	pdeo->closing = true;
ca469f35 AV	227	spin_unlock(&pde->pde_unload_lock);
ca469f35 AV	228	file = pdeo->file;
d56c0d45	229	pde->proc_ops->proc_release(file_inode(file), file);
ca469f35	230	spin_lock(&pde->pde_unload_lock);
492b2da6	231	/* After ->release. */
06a0c417	232	list_del(&pdeo->lh);
2f897424 AD	233	c = pdeo->c;
	234	spin_unlock(&pde->pde_unload_lock);
	235	if (unlikely(c))
	236	complete(c);
195b8cf0	237	kmem_cache_free(pde_opener_cache, pdeo);
05c0ae21	238	}
ca469f35 AV	239	}
ca469f35 AV	240
866ad9a7	241	void proc_entry_rundown(struct proc_dir_entry *de)
786d7e16	242	{
05c0ae21	243	DECLARE_COMPLETION_ONSTACK(c);
866ad9a7	244	/* Wait until all existing callers into module are done. */
05c0ae21 AV	245	de->pde_unload_completion = &c;
	246	if (atomic_add_return(BIAS, &de->in_use) != BIAS)
	247	wait_for_completion(&c);
786d7e16	248
492b2da6 AD	249	/* ->pde_openers list can't grow from now on. */
492b2da6 AD	250
05c0ae21	251	spin_lock(&de->pde_unload_lock);
866ad9a7 AV	252	while (!list_empty(&de->pde_openers)) {
866ad9a7 AV	253	struct pde_opener *pdeo;
866ad9a7	254	pdeo = list_first_entry(&de->pde_openers, struct pde_opener, lh);
ca469f35	255	close_pdeo(de, pdeo);
2f897424	256	spin_lock(&de->pde_unload_lock);
866ad9a7 AV	257	}
866ad9a7 AV	258	spin_unlock(&de->pde_unload_lock);
786d7e16 AD	259	}
786d7e16 AD	260
866ad9a7 AV	261	static loff_t proc_reg_llseek(struct file *file, loff_t offset, int whence)
	262	{
	263	struct proc_dir_entry *pde = PDE(file_inode(file));
	264	loff_t rv = -EINVAL;
	265	if (use_pde(pde)) {
d56c0d45	266	typeof_member(struct proc_ops, proc_lseek) lseek;
9af27b28	267
d56c0d45 AD	268	lseek = pde->proc_ops->proc_lseek;
	269	if (!lseek)
	270	lseek = default_llseek;
	271	rv = lseek(file, offset, whence);
866ad9a7 AV	272	unuse_pde(pde);
866ad9a7 AV	273	}
786d7e16 AD	274	return rv;
	275	}
	276
866ad9a7	277	static ssize_t proc_reg_read(struct file file, char __user buf, size_t count, loff_t *ppos)
786d7e16	278	{
496ad9aa	279	struct proc_dir_entry *pde = PDE(file_inode(file));
786d7e16	280	ssize_t rv = -EIO;
866ad9a7	281	if (use_pde(pde)) {
d56c0d45	282	typeof_member(struct proc_ops, proc_read) read;
9af27b28	283
d56c0d45	284	read = pde->proc_ops->proc_read;
866ad9a7 AV	285	if (read)
	286	rv = read(file, buf, count, ppos);
	287	unuse_pde(pde);
786d7e16	288	}
866ad9a7 AV	289	return rv;
866ad9a7 AV	290	}
786d7e16	291
866ad9a7 AV	292	static ssize_t proc_reg_write(struct file file, const char __user buf, size_t count, loff_t *ppos)
866ad9a7 AV	293	{
866ad9a7 AV	294	struct proc_dir_entry *pde = PDE(file_inode(file));
	295	ssize_t rv = -EIO;
	296	if (use_pde(pde)) {
d56c0d45	297	typeof_member(struct proc_ops, proc_write) write;
9af27b28	298
d56c0d45	299	write = pde->proc_ops->proc_write;
866ad9a7 AV	300	if (write)
	301	rv = write(file, buf, count, ppos);
	302	unuse_pde(pde);
	303	}
786d7e16 AD	304	return rv;
	305	}
	306
076ccb76	307	static __poll_t proc_reg_poll(struct file file, struct poll_table_struct pts)
786d7e16	308	{
496ad9aa	309	struct proc_dir_entry *pde = PDE(file_inode(file));
e6c8adca	310	__poll_t rv = DEFAULT_POLLMASK;
866ad9a7	311	if (use_pde(pde)) {
d56c0d45	312	typeof_member(struct proc_ops, proc_poll) poll;
9af27b28	313
d56c0d45	314	poll = pde->proc_ops->proc_poll;
866ad9a7 AV	315	if (poll)
	316	rv = poll(file, pts);
	317	unuse_pde(pde);
786d7e16	318	}
786d7e16 AD	319	return rv;
	320	}
	321
	322	static long proc_reg_unlocked_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
	323	{
496ad9aa	324	struct proc_dir_entry *pde = PDE(file_inode(file));
786d7e16	325	long rv = -ENOTTY;
866ad9a7	326	if (use_pde(pde)) {
d56c0d45	327	typeof_member(struct proc_ops, proc_ioctl) ioctl;
9af27b28	328
d56c0d45	329	ioctl = pde->proc_ops->proc_ioctl;
866ad9a7 AV	330	if (ioctl)
	331	rv = ioctl(file, cmd, arg);
	332	unuse_pde(pde);
786d7e16	333	}
786d7e16 AD	334	return rv;
	335	}
	336
	337	#ifdef CONFIG_COMPAT
	338	static long proc_reg_compat_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
	339	{
496ad9aa	340	struct proc_dir_entry *pde = PDE(file_inode(file));
786d7e16	341	long rv = -ENOTTY;
866ad9a7	342	if (use_pde(pde)) {
d56c0d45	343	typeof_member(struct proc_ops, proc_compat_ioctl) compat_ioctl;
9af27b28	344
d56c0d45	345	compat_ioctl = pde->proc_ops->proc_compat_ioctl;
866ad9a7 AV	346	if (compat_ioctl)
	347	rv = compat_ioctl(file, cmd, arg);
	348	unuse_pde(pde);
786d7e16	349	}
786d7e16 AD	350	return rv;
	351	}
	352	#endif
	353
	354	static int proc_reg_mmap(struct file file, struct vm_area_struct vma)
	355	{
496ad9aa	356	struct proc_dir_entry *pde = PDE(file_inode(file));
786d7e16	357	int rv = -EIO;
866ad9a7	358	if (use_pde(pde)) {
d56c0d45	359	typeof_member(struct proc_ops, proc_mmap) mmap;
9af27b28	360
d56c0d45	361	mmap = pde->proc_ops->proc_mmap;
866ad9a7 AV	362	if (mmap)
	363	rv = mmap(file, vma);
	364	unuse_pde(pde);
786d7e16	365	}
786d7e16 AD	366	return rv;
	367	}
	368
5721cf84 HD	369	static unsigned long
	370	proc_reg_get_unmapped_area(struct file *file, unsigned long orig_addr,
	371	unsigned long len, unsigned long pgoff,
	372	unsigned long flags)
c4fe2448 AD	373	{
c4fe2448 AD	374	struct proc_dir_entry *pde = PDE(file_inode(file));
2cbe3b0a	375	unsigned long rv = -EIO;
ae5758a1	376
c4fe2448	377	if (use_pde(pde)) {
d56c0d45	378	typeof_member(struct proc_ops, proc_get_unmapped_area) get_area;
ae5758a1	379
d56c0d45	380	get_area = pde->proc_ops->proc_get_unmapped_area;
fad1a86e	381	#ifdef CONFIG_MMU
ae5758a1 JB	382	if (!get_area)
ae5758a1 JB	383	get_area = current->mm->get_unmapped_area;
fad1a86e	384	#endif
ae5758a1	385
5721cf84 HD	386	if (get_area)
5721cf84 HD	387	rv = get_area(file, orig_addr, len, pgoff, flags);
ae5758a1 JB	388	else
ae5758a1 JB	389	rv = orig_addr;
c4fe2448 AD	390	unuse_pde(pde);
	391	}
	392	return rv;
	393	}
	394
786d7e16 AD	395	static int proc_reg_open(struct inode inode, struct file file)
	396	{
	397	struct proc_dir_entry *pde = PDE(inode);
	398	int rv = 0;
d56c0d45 AD	399	typeof_member(struct proc_ops, proc_open) open;
d56c0d45 AD	400	typeof_member(struct proc_ops, proc_release) release;
881adb85 AD	401	struct pde_opener *pdeo;
	402
	403	/*
492b2da6 AD	404	* Ensure that
	405	* 1) PDE's ->release hook will be called no matter what
	406	* either normally by close()/->release, or forcefully by
	407	* rmmod/remove_proc_entry.
	408	*
	409	* 2) rmmod isn't blocked by opening file in /proc and sitting on
	410	* the descriptor (including "rmmod foo </proc/foo" scenario).
881adb85	411	*
492b2da6	412	* Save every "struct file" with custom ->release hook.
881adb85	413	*/
e7a6e291	414	if (!use_pde(pde))
d2857e79	415	return -ENOENT;
e7a6e291	416
d56c0d45	417	release = pde->proc_ops->proc_release;
e7a6e291	418	if (release) {
195b8cf0	419	pdeo = kmem_cache_alloc(pde_opener_cache, GFP_KERNEL);
e7a6e291 AD	420	if (!pdeo) {
	421	rv = -ENOMEM;
	422	goto out_unuse;
	423	}
	424	}
786d7e16	425
d56c0d45	426	open = pde->proc_ops->proc_open;
786d7e16 AD	427	if (open)
	428	rv = open(inode, file);
	429
e7a6e291 AD	430	if (release) {
	431	if (rv == 0) {
	432	/* To know what to release. */
	433	pdeo->file = file;
	434	pdeo->closing = false;
	435	pdeo->c = NULL;
	436	spin_lock(&pde->pde_unload_lock);
	437	list_add(&pdeo->lh, &pde->pde_openers);
	438	spin_unlock(&pde->pde_unload_lock);
	439	} else
195b8cf0	440	kmem_cache_free(pde_opener_cache, pdeo);
e7a6e291	441	}
05c0ae21	442
e7a6e291	443	out_unuse:
05c0ae21	444	unuse_pde(pde);
786d7e16 AD	445	return rv;
	446	}
	447
	448	static int proc_reg_release(struct inode inode, struct file file)
	449	{
	450	struct proc_dir_entry *pde = PDE(inode);
881adb85	451	struct pde_opener *pdeo;
786d7e16	452	spin_lock(&pde->pde_unload_lock);
ca469f35 AV	453	list_for_each_entry(pdeo, &pde->pde_openers, lh) {
	454	if (pdeo->file == file) {
	455	close_pdeo(pde, pdeo);
2f897424	456	return 0;
ca469f35	457	}
881adb85	458	}
786d7e16	459	spin_unlock(&pde->pde_unload_lock);
ca469f35	460	return 0;
786d7e16 AD	461	}
	462
	463	static const struct file_operations proc_reg_file_ops = {
	464	.llseek = proc_reg_llseek,
	465	.read = proc_reg_read,
	466	.write = proc_reg_write,
	467	.poll = proc_reg_poll,
	468	.unlocked_ioctl = proc_reg_unlocked_ioctl,
	469	#ifdef CONFIG_COMPAT
	470	.compat_ioctl = proc_reg_compat_ioctl,
	471	#endif
	472	.mmap = proc_reg_mmap,
c4fe2448	473	.get_unmapped_area = proc_reg_get_unmapped_area,
786d7e16 AD	474	.open = proc_reg_open,
	475	.release = proc_reg_release,
	476	};
	477
778f3dd5 DM	478	#ifdef CONFIG_COMPAT
	479	static const struct file_operations proc_reg_file_ops_no_compat = {
	480	.llseek = proc_reg_llseek,
	481	.read = proc_reg_read,
	482	.write = proc_reg_write,
	483	.poll = proc_reg_poll,
	484	.unlocked_ioctl = proc_reg_unlocked_ioctl,
	485	.mmap = proc_reg_mmap,
c4fe2448	486	.get_unmapped_area = proc_reg_get_unmapped_area,
778f3dd5 DM	487	.open = proc_reg_open,
	488	.release = proc_reg_release,
	489	};
	490	#endif
	491
fceef393 AV	492	static void proc_put_link(void *p)
	493	{
	494	unuse_pde(p);
	495	}
	496
6b255391	497	static const char proc_get_link(struct dentry dentry,
fceef393 AV	498	struct inode *inode,
fceef393 AV	499	struct delayed_call *done)
7e0e953b	500	{
6b255391	501	struct proc_dir_entry *pde = PDE(inode);
15b158b4	502	if (!use_pde(pde))
7e0e953b	503	return ERR_PTR(-EINVAL);
fceef393	504	set_delayed_call(done, proc_put_link, pde);
680baacb	505	return pde->data;
7e0e953b AV	506	}
7e0e953b AV	507
7e0e953b	508	const struct inode_operations proc_link_inode_operations = {
6b255391	509	.get_link = proc_get_link,
7e0e953b AV	510	};
7e0e953b AV	511
6d1b6e4e	512	struct inode proc_get_inode(struct super_block sb, struct proc_dir_entry *de)
1da177e4	513	{
51f0885e	514	struct inode *inode = new_inode_pseudo(sb);
1da177e4	515
51f0885e LT	516	if (inode) {
51f0885e LT	517	inode->i_ino = de->low_ino;
078cd827	518	inode->i_mtime = inode->i_atime = inode->i_ctime = current_time(inode);
a1d4aebb	519	PROC_I(inode)->pde = de;
5e971dce	520
eb6d38d5 EB	521	if (is_empty_pde(de)) {
	522	make_empty_dir_inode(inode);
	523	return inode;
	524	}
5e971dce AD	525	if (de->mode) {
	526	inode->i_mode = de->mode;
	527	inode->i_uid = de->uid;
	528	inode->i_gid = de->gid;
	529	}
	530	if (de->size)
	531	inode->i_size = de->size;
	532	if (de->nlink)
bfe86848	533	set_nlink(inode, de->nlink);
d56c0d45 AD	534
	535	if (S_ISREG(inode->i_mode)) {
	536	inode->i_op = de->proc_iops;
	537	inode->i_fop = &proc_reg_file_ops;
778f3dd5	538	#ifdef CONFIG_COMPAT
d56c0d45 AD	539	if (!de->proc_ops->proc_compat_ioctl) {
d56c0d45 AD	540	inode->i_fop = &proc_reg_file_ops_no_compat;
778f3dd5	541	}
d56c0d45 AD	542	#endif
	543	} else if (S_ISDIR(inode->i_mode)) {
	544	inode->i_op = de->proc_iops;
	545	inode->i_fop = de->proc_dir_ops;
	546	} else if (S_ISLNK(inode->i_mode)) {
	547	inode->i_op = de->proc_iops;
	548	inode->i_fop = NULL;
	549	} else
	550	BUG();
99b76233	551	} else
135d5655	552	pde_put(de);
1da177e4	553	return inode;
d3d009cb	554	}