[mirror_ubuntu-focal-kernel.git] / fs / proc / inode.c

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

#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/magic.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;

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

	/* Stop tracking associated processes */
	put_pid(PROC_I(inode)->pid);

	/* Let go of any associated proc directory entry */
	de = PDE(inode);
	if (de)
		pde_put(de);

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

static struct kmem_cache * proc_inode_cachep;

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

	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;
	ei->ns_ops = NULL;
	inode = &ei->vfs_inode;
	return inode;
}

static void proc_i_callback(struct rcu_head *head)
{
	struct inode *inode = container_of(head, struct inode, i_rcu);
	kmem_cache_free(proc_inode_cachep, PROC_I(inode));
}

static void proc_destroy_inode(struct inode *inode)
{
	call_rcu(&inode->i_rcu, proc_i_callback);
}

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

	inode_init_once(&ei->vfs_inode);
}

void __init proc_init_inodecache(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);
}

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;
}

static const struct super_operations proc_sops = {
	.alloc_inode	= proc_alloc_inode,
	.destroy_inode	= proc_destroy_inode,
	.drop_inode	= generic_delete_inode,
	.evict_inode	= proc_evict_inode,
	.statfs		= simple_statfs,
	.remount_fs	= proc_remount,
	.show_options	= proc_show_options,
};

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

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

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

/* pde is locked */
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);
		spin_lock(&pde->pde_unload_lock);
	} else {
		struct file *file;
		pdeo->closing = true;
		spin_unlock(&pde->pde_unload_lock);
		file = pdeo->file;
		pde->proc_fops->release(file_inode(file), file);
		spin_lock(&pde->pde_unload_lock);
		/* After ->release. */
		list_del(&pdeo->lh);
		if (pdeo->c)
			complete(pdeo->c);
		kfree(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_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)) {
		loff_t (*llseek)(struct file *, loff_t, int);
		llseek = pde->proc_fops->llseek;
		if (!llseek)
			llseek = default_llseek;
		rv = llseek(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)
{
	ssize_t (*read)(struct file *, char __user *, size_t, loff_t *);
	struct proc_dir_entry *pde = PDE(file_inode(file));
	ssize_t rv = -EIO;
	if (use_pde(pde)) {
		read = pde->proc_fops->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)
{
	ssize_t (*write)(struct file *, const char __user *, size_t, loff_t *);
	struct proc_dir_entry *pde = PDE(file_inode(file));
	ssize_t rv = -EIO;
	if (use_pde(pde)) {
		write = pde->proc_fops->write;
		if (write)
			rv = write(file, buf, count, ppos);
		unuse_pde(pde);
	}
	return rv;
}

static unsigned int proc_reg_poll(struct file *file, struct poll_table_struct *pts)
{
	struct proc_dir_entry *pde = PDE(file_inode(file));
	unsigned int rv = DEFAULT_POLLMASK;
	unsigned int (*poll)(struct file *, struct poll_table_struct *);
	if (use_pde(pde)) {
		poll = pde->proc_fops->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;
	long (*ioctl)(struct file *, unsigned int, unsigned long);
	if (use_pde(pde)) {
		ioctl = pde->proc_fops->unlocked_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;
	long (*compat_ioctl)(struct file *, unsigned int, unsigned long);
	if (use_pde(pde)) {
		compat_ioctl = pde->proc_fops->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;
	int (*mmap)(struct file *, struct vm_area_struct *);
	if (use_pde(pde)) {
		mmap = pde->proc_fops->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(proc_reg_get_unmapped_area) *get_area;

		get_area = pde->proc_fops->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;
	int (*open)(struct inode *, struct file *);
	int (*release)(struct inode *, struct file *);
	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.
	 */
	pdeo = kmalloc(sizeof(struct pde_opener), GFP_KERNEL);
	if (!pdeo)
		return -ENOMEM;

	if (!use_pde(pde)) {
		kfree(pdeo);
		return -ENOENT;
	}
	open = pde->proc_fops->open;
	release = pde->proc_fops->release;

	if (open)
		rv = open(inode, file);

	if (rv == 0 && release) {
		/* 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
		kfree(pdeo);

	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);
			break;
		}
	}
	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 (unlikely(!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);
		WARN_ON(!de->proc_iops);
		inode->i_op = de->proc_iops;
		if (de->proc_fops) {
			if (S_ISREG(inode->i_mode)) {
#ifdef CONFIG_COMPAT
				if (!de->proc_fops->compat_ioctl)
					inode->i_fop =
						&proc_reg_file_ops_no_compat;
				else
#endif
					inode->i_fop = &proc_reg_file_ops;
			} else {
				inode->i_fop = de->proc_fops;
			}
		}
	} else
	       pde_put(de);
	return inode;
}

int proc_fill_super(struct super_block *s, void *data, int silent)
{
	struct pid_namespace *ns = get_pid_ns(s->s_fs_info);
	struct inode *root_inode;
	int ret;

	if (!proc_parse_options(data, ns))
		return -EINVAL;

	/* 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;
	
	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);
}
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
	8	#include <linux/time.h>
	9	#include <linux/proc_fs.h>
	10	#include <linux/kernel.h>
97412950	11	#include <linux/pid_namespace.h>
1da177e4 LT	12	#include <linux/mm.h>
	13	#include <linux/string.h>
	14	#include <linux/stat.h>
786d7e16	15	#include <linux/completion.h>
dd23aae4	16	#include <linux/poll.h>
87ebdc00	17	#include <linux/printk.h>
1da177e4 LT	18	#include <linux/file.h>
	19	#include <linux/limits.h>
	20	#include <linux/init.h>
	21	#include <linux/module.h>
9043476f	22	#include <linux/sysctl.h>
97412950	23	#include <linux/seq_file.h>
5a0e3ad6	24	#include <linux/slab.h>
97412950	25	#include <linux/mount.h>
303eb7e2	26	#include <linux/magic.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;
1da177e4	36
91b0abe3	37	truncate_inode_pages_final(&inode->i_data);
dbd5768f	38	clear_inode(inode);
fef26658	39
99f89551	40	/* Stop tracking associated processes */
13b41b09	41	put_pid(PROC_I(inode)->pid);
1da177e4 LT	42
1da177e4 LT	43	/* Let go of any associated proc directory entry */
6bee55f9	44	de = PDE(inode);
99b76233	45	if (de)
135d5655	46	pde_put(de);
d6cffbbe	47
dfef6dcd AV	48	head = PROC_I(inode)->sysctl;
dfef6dcd AV	49	if (head) {
1c44dbc8	50	RCU_INIT_POINTER(PROC_I(inode)->sysctl, NULL);
d6cffbbe	51	proc_sys_evict_inode(inode, head);
dfef6dcd	52	}
1da177e4 LT	53	}
1da177e4 LT	54
e18b890b	55	static struct kmem_cache * proc_inode_cachep;
1da177e4 LT	56
	57	static struct inode proc_alloc_inode(struct super_block sb)
	58	{
	59	struct proc_inode *ei;
	60	struct inode *inode;
	61
f245e1c1	62	ei = kmem_cache_alloc(proc_inode_cachep, GFP_KERNEL);
1da177e4 LT	63	if (!ei)
1da177e4 LT	64	return NULL;
13b41b09	65	ei->pid = NULL;
aed7a6c4	66	ei->fd = 0;
1da177e4 LT	67	ei->op.proc_get_link = NULL;
1da177e4 LT	68	ei->pde = NULL;
9043476f AV	69	ei->sysctl = NULL;
9043476f AV	70	ei->sysctl_entry = NULL;
3d3d35b1	71	ei->ns_ops = NULL;
1da177e4	72	inode = &ei->vfs_inode;
1da177e4 LT	73	return inode;
	74	}
	75
fa0d7e3d	76	static void proc_i_callback(struct rcu_head *head)
1da177e4	77	{
fa0d7e3d	78	struct inode *inode = container_of(head, struct inode, i_rcu);
1da177e4 LT	79	kmem_cache_free(proc_inode_cachep, PROC_I(inode));
	80	}
	81
fa0d7e3d NP	82	static void proc_destroy_inode(struct inode *inode)
	83	{
	84	call_rcu(&inode->i_rcu, proc_i_callback);
	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
5bcd7ff9	94	void __init proc_init_inodecache(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);
1da177e4 LT	102	}
1da177e4 LT	103
97412950 VK	104	static int proc_show_options(struct seq_file seq, struct dentry root)
97412950 VK	105	{
0499680a VK	106	struct super_block *sb = root->d_sb;
	107	struct pid_namespace *pid = sb->s_fs_info;
	108
dcb0f222 EB	109	if (!gid_eq(pid->pid_gid, GLOBAL_ROOT_GID))
dcb0f222 EB	110	seq_printf(seq, ",gid=%u", from_kgid_munged(&init_user_ns, pid->pid_gid));
796f571b	111	if (pid->hide_pid != HIDEPID_OFF)
0499680a VK	112	seq_printf(seq, ",hidepid=%u", pid->hide_pid);
0499680a VK	113
97412950 VK	114	return 0;
	115	}
	116
ee9b6d61	117	static const struct super_operations proc_sops = {
1da177e4 LT	118	.alloc_inode = proc_alloc_inode,
1da177e4 LT	119	.destroy_inode = proc_destroy_inode,
1da177e4	120	.drop_inode = generic_delete_inode,
8267952b	121	.evict_inode = proc_evict_inode,
1da177e4	122	.statfs = simple_statfs,
97412950 VK	123	.remount_fs = proc_remount,
97412950 VK	124	.show_options = proc_show_options,
1da177e4 LT	125	};
1da177e4 LT	126
866ad9a7 AV	127	enum {BIAS = -1U<<31};
	128
	129	static inline int use_pde(struct proc_dir_entry *pde)
	130	{
05c0ae21	131	return atomic_inc_unless_negative(&pde->in_use);
881adb85 AD	132	}
881adb85 AD	133
866ad9a7	134	static void unuse_pde(struct proc_dir_entry *pde)
881adb85	135	{
05c0ae21 AV	136	if (atomic_dec_return(&pde->in_use) == BIAS)
05c0ae21 AV	137	complete(pde->pde_unload_completion);
786d7e16 AD	138	}
786d7e16 AD	139
ca469f35 AV	140	/* pde is locked */
	141	static void close_pdeo(struct proc_dir_entry pde, struct pde_opener pdeo)
	142	{
492b2da6 AD	143	/*
	144	* close() (proc_reg_release()) can't delete an entry and proceed:
	145	* ->release hook needs to be available at the right moment.
	146	*
	147	* rmmod (remove_proc_entry() et al) can't delete an entry and proceed:
	148	* "struct file" needs to be available at the right moment.
	149	*
	150	* Therefore, first process to enter this function does ->release() and
	151	* signals its completion to the other process which does nothing.
	152	*/
05c0ae21	153	if (pdeo->closing) {
ca469f35	154	/* somebody else is doing that, just wait */
05c0ae21 AV	155	DECLARE_COMPLETION_ONSTACK(c);
05c0ae21 AV	156	pdeo->c = &c;
ca469f35	157	spin_unlock(&pde->pde_unload_lock);
05c0ae21	158	wait_for_completion(&c);
ca469f35	159	spin_lock(&pde->pde_unload_lock);
ca469f35 AV	160	} else {
ca469f35 AV	161	struct file *file;
f5887c71	162	pdeo->closing = true;
ca469f35 AV	163	spin_unlock(&pde->pde_unload_lock);
	164	file = pdeo->file;
	165	pde->proc_fops->release(file_inode(file), file);
	166	spin_lock(&pde->pde_unload_lock);
492b2da6	167	/* After ->release. */
06a0c417	168	list_del(&pdeo->lh);
05c0ae21 AV	169	if (pdeo->c)
05c0ae21 AV	170	complete(pdeo->c);
ca469f35	171	kfree(pdeo);
05c0ae21	172	}
ca469f35 AV	173	}
ca469f35 AV	174
866ad9a7	175	void proc_entry_rundown(struct proc_dir_entry *de)
786d7e16	176	{
05c0ae21	177	DECLARE_COMPLETION_ONSTACK(c);
866ad9a7	178	/* Wait until all existing callers into module are done. */
05c0ae21 AV	179	de->pde_unload_completion = &c;
	180	if (atomic_add_return(BIAS, &de->in_use) != BIAS)
	181	wait_for_completion(&c);
786d7e16	182
492b2da6 AD	183	/* ->pde_openers list can't grow from now on. */
492b2da6 AD	184
05c0ae21	185	spin_lock(&de->pde_unload_lock);
866ad9a7 AV	186	while (!list_empty(&de->pde_openers)) {
866ad9a7 AV	187	struct pde_opener *pdeo;
866ad9a7	188	pdeo = list_first_entry(&de->pde_openers, struct pde_opener, lh);
ca469f35	189	close_pdeo(de, pdeo);
866ad9a7 AV	190	}
866ad9a7 AV	191	spin_unlock(&de->pde_unload_lock);
786d7e16 AD	192	}
786d7e16 AD	193
866ad9a7 AV	194	static loff_t proc_reg_llseek(struct file *file, loff_t offset, int whence)
	195	{
	196	struct proc_dir_entry *pde = PDE(file_inode(file));
	197	loff_t rv = -EINVAL;
	198	if (use_pde(pde)) {
	199	loff_t (llseek)(struct file , loff_t, int);
	200	llseek = pde->proc_fops->llseek;
	201	if (!llseek)
	202	llseek = default_llseek;
	203	rv = llseek(file, offset, whence);
	204	unuse_pde(pde);
	205	}
786d7e16 AD	206	return rv;
	207	}
	208
866ad9a7	209	static ssize_t proc_reg_read(struct file file, char __user buf, size_t count, loff_t *ppos)
786d7e16	210	{
866ad9a7	211	ssize_t (read)(struct file , char __user , size_t, loff_t );
496ad9aa	212	struct proc_dir_entry *pde = PDE(file_inode(file));
786d7e16	213	ssize_t rv = -EIO;
866ad9a7 AV	214	if (use_pde(pde)) {
	215	read = pde->proc_fops->read;
	216	if (read)
	217	rv = read(file, buf, count, ppos);
	218	unuse_pde(pde);
786d7e16	219	}
866ad9a7 AV	220	return rv;
866ad9a7 AV	221	}
786d7e16	222
866ad9a7 AV	223	static ssize_t proc_reg_write(struct file file, const char __user buf, size_t count, loff_t *ppos)
	224	{
	225	ssize_t (write)(struct file , const char __user , size_t, loff_t );
	226	struct proc_dir_entry *pde = PDE(file_inode(file));
	227	ssize_t rv = -EIO;
	228	if (use_pde(pde)) {
	229	write = pde->proc_fops->write;
	230	if (write)
	231	rv = write(file, buf, count, ppos);
	232	unuse_pde(pde);
	233	}
786d7e16 AD	234	return rv;
	235	}
	236
	237	static unsigned int proc_reg_poll(struct file file, struct poll_table_struct pts)
	238	{
496ad9aa	239	struct proc_dir_entry *pde = PDE(file_inode(file));
dd23aae4	240	unsigned int rv = DEFAULT_POLLMASK;
786d7e16	241	unsigned int (poll)(struct file , struct poll_table_struct *);
866ad9a7 AV	242	if (use_pde(pde)) {
	243	poll = pde->proc_fops->poll;
	244	if (poll)
	245	rv = poll(file, pts);
	246	unuse_pde(pde);
786d7e16	247	}
786d7e16 AD	248	return rv;
	249	}
	250
	251	static long proc_reg_unlocked_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
	252	{
496ad9aa	253	struct proc_dir_entry *pde = PDE(file_inode(file));
786d7e16	254	long rv = -ENOTTY;
b19dd42f	255	long (ioctl)(struct file , unsigned int, unsigned long);
866ad9a7 AV	256	if (use_pde(pde)) {
	257	ioctl = pde->proc_fops->unlocked_ioctl;
	258	if (ioctl)
	259	rv = ioctl(file, cmd, arg);
	260	unuse_pde(pde);
786d7e16	261	}
786d7e16 AD	262	return rv;
	263	}
	264
	265	#ifdef CONFIG_COMPAT
	266	static long proc_reg_compat_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
	267	{
496ad9aa	268	struct proc_dir_entry *pde = PDE(file_inode(file));
786d7e16 AD	269	long rv = -ENOTTY;
786d7e16 AD	270	long (compat_ioctl)(struct file , unsigned int, unsigned long);
866ad9a7 AV	271	if (use_pde(pde)) {
	272	compat_ioctl = pde->proc_fops->compat_ioctl;
	273	if (compat_ioctl)
	274	rv = compat_ioctl(file, cmd, arg);
	275	unuse_pde(pde);
786d7e16	276	}
786d7e16 AD	277	return rv;
	278	}
	279	#endif
	280
	281	static int proc_reg_mmap(struct file file, struct vm_area_struct vma)
	282	{
496ad9aa	283	struct proc_dir_entry *pde = PDE(file_inode(file));
786d7e16 AD	284	int rv = -EIO;
786d7e16 AD	285	int (mmap)(struct file , struct vm_area_struct *);
866ad9a7 AV	286	if (use_pde(pde)) {
	287	mmap = pde->proc_fops->mmap;
	288	if (mmap)
	289	rv = mmap(file, vma);
	290	unuse_pde(pde);
786d7e16	291	}
786d7e16 AD	292	return rv;
	293	}
	294
5721cf84 HD	295	static unsigned long
	296	proc_reg_get_unmapped_area(struct file *file, unsigned long orig_addr,
	297	unsigned long len, unsigned long pgoff,
	298	unsigned long flags)
c4fe2448 AD	299	{
c4fe2448 AD	300	struct proc_dir_entry *pde = PDE(file_inode(file));
2cbe3b0a	301	unsigned long rv = -EIO;
ae5758a1	302
c4fe2448	303	if (use_pde(pde)) {
ae5758a1 JB	304	typeof(proc_reg_get_unmapped_area) *get_area;
	305
	306	get_area = pde->proc_fops->get_unmapped_area;
fad1a86e	307	#ifdef CONFIG_MMU
ae5758a1 JB	308	if (!get_area)
ae5758a1 JB	309	get_area = current->mm->get_unmapped_area;
fad1a86e	310	#endif
ae5758a1	311
5721cf84 HD	312	if (get_area)
5721cf84 HD	313	rv = get_area(file, orig_addr, len, pgoff, flags);
ae5758a1 JB	314	else
ae5758a1 JB	315	rv = orig_addr;
c4fe2448 AD	316	unuse_pde(pde);
	317	}
	318	return rv;
	319	}
	320
786d7e16 AD	321	static int proc_reg_open(struct inode inode, struct file file)
	322	{
	323	struct proc_dir_entry *pde = PDE(inode);
	324	int rv = 0;
	325	int (open)(struct inode , struct file *);
881adb85 AD	326	int (release)(struct inode , struct file *);
	327	struct pde_opener *pdeo;
	328
	329	/*
492b2da6 AD	330	* Ensure that
	331	* 1) PDE's ->release hook will be called no matter what
	332	* either normally by close()/->release, or forcefully by
	333	* rmmod/remove_proc_entry.
	334	*
	335	* 2) rmmod isn't blocked by opening file in /proc and sitting on
	336	* the descriptor (including "rmmod foo </proc/foo" scenario).
881adb85	337	*
492b2da6	338	* Save every "struct file" with custom ->release hook.
881adb85	339	*/
39a10ac2	340	pdeo = kmalloc(sizeof(struct pde_opener), GFP_KERNEL);
881adb85 AD	341	if (!pdeo)
881adb85 AD	342	return -ENOMEM;
786d7e16	343
866ad9a7	344	if (!use_pde(pde)) {
881adb85	345	kfree(pdeo);
d2857e79	346	return -ENOENT;
786d7e16	347	}
786d7e16	348	open = pde->proc_fops->open;
881adb85	349	release = pde->proc_fops->release;
786d7e16 AD	350
	351	if (open)
	352	rv = open(inode, file);
	353
881adb85 AD	354	if (rv == 0 && release) {
881adb85 AD	355	/* To know what to release. */
881adb85	356	pdeo->file = file;
39a10ac2 AD	357	pdeo->closing = false;
39a10ac2 AD	358	pdeo->c = NULL;
05c0ae21	359	spin_lock(&pde->pde_unload_lock);
881adb85	360	list_add(&pdeo->lh, &pde->pde_openers);
05c0ae21	361	spin_unlock(&pde->pde_unload_lock);
881adb85 AD	362	} else
881adb85 AD	363	kfree(pdeo);
05c0ae21 AV	364
05c0ae21 AV	365	unuse_pde(pde);
786d7e16 AD	366	return rv;
	367	}
	368
	369	static int proc_reg_release(struct inode inode, struct file file)
	370	{
	371	struct proc_dir_entry *pde = PDE(inode);
881adb85	372	struct pde_opener *pdeo;
786d7e16	373	spin_lock(&pde->pde_unload_lock);
ca469f35 AV	374	list_for_each_entry(pdeo, &pde->pde_openers, lh) {
	375	if (pdeo->file == file) {
	376	close_pdeo(pde, pdeo);
	377	break;
	378	}
881adb85	379	}
786d7e16	380	spin_unlock(&pde->pde_unload_lock);
ca469f35	381	return 0;
786d7e16 AD	382	}
	383
	384	static const struct file_operations proc_reg_file_ops = {
	385	.llseek = proc_reg_llseek,
	386	.read = proc_reg_read,
	387	.write = proc_reg_write,
	388	.poll = proc_reg_poll,
	389	.unlocked_ioctl = proc_reg_unlocked_ioctl,
	390	#ifdef CONFIG_COMPAT
	391	.compat_ioctl = proc_reg_compat_ioctl,
	392	#endif
	393	.mmap = proc_reg_mmap,
c4fe2448	394	.get_unmapped_area = proc_reg_get_unmapped_area,
786d7e16 AD	395	.open = proc_reg_open,
	396	.release = proc_reg_release,
	397	};
	398
778f3dd5 DM	399	#ifdef CONFIG_COMPAT
	400	static const struct file_operations proc_reg_file_ops_no_compat = {
	401	.llseek = proc_reg_llseek,
	402	.read = proc_reg_read,
	403	.write = proc_reg_write,
	404	.poll = proc_reg_poll,
	405	.unlocked_ioctl = proc_reg_unlocked_ioctl,
	406	.mmap = proc_reg_mmap,
c4fe2448	407	.get_unmapped_area = proc_reg_get_unmapped_area,
778f3dd5 DM	408	.open = proc_reg_open,
	409	.release = proc_reg_release,
	410	};
	411	#endif
	412
fceef393 AV	413	static void proc_put_link(void *p)
	414	{
	415	unuse_pde(p);
	416	}
	417
6b255391	418	static const char proc_get_link(struct dentry dentry,
fceef393 AV	419	struct inode *inode,
fceef393 AV	420	struct delayed_call *done)
7e0e953b	421	{
6b255391	422	struct proc_dir_entry *pde = PDE(inode);
7e0e953b AV	423	if (unlikely(!use_pde(pde)))
7e0e953b AV	424	return ERR_PTR(-EINVAL);
fceef393	425	set_delayed_call(done, proc_put_link, pde);
680baacb	426	return pde->data;
7e0e953b AV	427	}
7e0e953b AV	428
7e0e953b	429	const struct inode_operations proc_link_inode_operations = {
6b255391	430	.get_link = proc_get_link,
7e0e953b AV	431	};
7e0e953b AV	432
6d1b6e4e	433	struct inode proc_get_inode(struct super_block sb, struct proc_dir_entry *de)
1da177e4	434	{
51f0885e	435	struct inode *inode = new_inode_pseudo(sb);
1da177e4	436
51f0885e LT	437	if (inode) {
51f0885e LT	438	inode->i_ino = de->low_ino;
078cd827	439	inode->i_mtime = inode->i_atime = inode->i_ctime = current_time(inode);
a1d4aebb	440	PROC_I(inode)->pde = de;
5e971dce	441
eb6d38d5 EB	442	if (is_empty_pde(de)) {
	443	make_empty_dir_inode(inode);
	444	return inode;
	445	}
5e971dce AD	446	if (de->mode) {
	447	inode->i_mode = de->mode;
	448	inode->i_uid = de->uid;
	449	inode->i_gid = de->gid;
	450	}
	451	if (de->size)
	452	inode->i_size = de->size;
	453	if (de->nlink)
bfe86848	454	set_nlink(inode, de->nlink);
b6cdc731 AV	455	WARN_ON(!de->proc_iops);
b6cdc731 AV	456	inode->i_op = de->proc_iops;
5e971dce AD	457	if (de->proc_fops) {
5e971dce AD	458	if (S_ISREG(inode->i_mode)) {
778f3dd5	459	#ifdef CONFIG_COMPAT
5e971dce AD	460	if (!de->proc_fops->compat_ioctl)
	461	inode->i_fop =
	462	&proc_reg_file_ops_no_compat;
	463	else
778f3dd5	464	#endif
5e971dce AD	465	inode->i_fop = &proc_reg_file_ops;
	466	} else {
	467	inode->i_fop = de->proc_fops;
778f3dd5	468	}
786d7e16	469	}
99b76233	470	} else
135d5655	471	pde_put(de);
1da177e4	472	return inode;
d3d009cb	473	}
1da177e4	474
e94591d0	475	int proc_fill_super(struct super_block s, void data, int silent)
1da177e4	476	{
e94591d0	477	struct pid_namespace *ns = get_pid_ns(s->s_fs_info);
87e0aab3	478	struct inode *root_inode;
0097875b	479	int ret;
87e0aab3	480
e94591d0 EB	481	if (!proc_parse_options(data, ns))
	482	return -EINVAL;
	483
a2982cc9 EB	484	/* User space would break if executables or devices appear on proc */
a2982cc9 EB	485	s->s_iflags \|= SB_I_USERNS_VISIBLE \| SB_I_NOEXEC \| SB_I_NODEV;
1751e8a6	486	s->s_flags \|= SB_NODIRATIME \| SB_NOSUID \| SB_NOEXEC;
1da177e4 LT	487	s->s_blocksize = 1024;
	488	s->s_blocksize_bits = 10;
	489	s->s_magic = PROC_SUPER_MAGIC;
	490	s->s_op = &proc_sops;
	491	s->s_time_gran = 1;
a867d734 LT	492
	493	/*
	494	* procfs isn't actually a stacking filesystem; however, there is
	495	* too much magic going on inside it to permit stacking things on
	496	* top of it
	497	*/
	498	s->s_stack_depth = FILESYSTEM_MAX_STACK_DEPTH;
1da177e4	499
135d5655	500	pde_get(&proc_root);
87e0aab3 MP	501	root_inode = proc_get_inode(s, &proc_root);
87e0aab3 MP	502	if (!root_inode) {
87ebdc00	503	pr_err("proc_fill_super: get root inode failed\n");
87e0aab3 MP	504	return -ENOMEM;
87e0aab3 MP	505	}
1da177e4	506
87e0aab3 MP	507	s->s_root = d_make_root(root_inode);
87e0aab3 MP	508	if (!s->s_root) {
87ebdc00	509	pr_err("proc_fill_super: allocate dentry failed\n");
87e0aab3 MP	510	return -ENOMEM;
	511	}
	512
0097875b EB	513	ret = proc_setup_self(s);
	514	if (ret) {
	515	return ret;
	516	}
	517	return proc_setup_thread_self(s);
1da177e4	518	}