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

/*
 *  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 |= MS_NODIRATIME | MS_NOSUID | MS_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
1da177e4 LT	1	/*
	2	* linux/fs/proc/inode.c
	3	*
	4	* Copyright (C) 1991, 1992 Linus Torvalds
	5	*/
	6
	7	#include <linux/time.h>
	8	#include <linux/proc_fs.h>
	9	#include <linux/kernel.h>
97412950	10	#include <linux/pid_namespace.h>
1da177e4 LT	11	#include <linux/mm.h>
	12	#include <linux/string.h>
	13	#include <linux/stat.h>
786d7e16	14	#include <linux/completion.h>
dd23aae4	15	#include <linux/poll.h>
87ebdc00	16	#include <linux/printk.h>
1da177e4 LT	17	#include <linux/file.h>
	18	#include <linux/limits.h>
	19	#include <linux/init.h>
	20	#include <linux/module.h>
9043476f	21	#include <linux/sysctl.h>
97412950	22	#include <linux/seq_file.h>
5a0e3ad6	23	#include <linux/slab.h>
97412950	24	#include <linux/mount.h>
303eb7e2	25	#include <linux/magic.h>
1da177e4	26
7c0f6ba6	27	#include <linux/uaccess.h>
1da177e4	28
fee781e6	29	#include "internal.h"
1da177e4	30
8267952b	31	static void proc_evict_inode(struct inode *inode)
1da177e4 LT	32	{
1da177e4 LT	33	struct proc_dir_entry *de;
dfef6dcd	34	struct ctl_table_header *head;
1da177e4	35
91b0abe3	36	truncate_inode_pages_final(&inode->i_data);
dbd5768f	37	clear_inode(inode);
fef26658	38
99f89551	39	/* Stop tracking associated processes */
13b41b09	40	put_pid(PROC_I(inode)->pid);
1da177e4 LT	41
1da177e4 LT	42	/* Let go of any associated proc directory entry */
6bee55f9	43	de = PDE(inode);
99b76233	44	if (de)
135d5655	45	pde_put(de);
d6cffbbe	46
dfef6dcd AV	47	head = PROC_I(inode)->sysctl;
dfef6dcd AV	48	if (head) {
1c44dbc8	49	RCU_INIT_POINTER(PROC_I(inode)->sysctl, NULL);
d6cffbbe	50	proc_sys_evict_inode(inode, head);
dfef6dcd	51	}
1da177e4 LT	52	}
1da177e4 LT	53
e18b890b	54	static struct kmem_cache * proc_inode_cachep;
1da177e4 LT	55
	56	static struct inode proc_alloc_inode(struct super_block sb)
	57	{
	58	struct proc_inode *ei;
	59	struct inode *inode;
	60
f245e1c1	61	ei = kmem_cache_alloc(proc_inode_cachep, GFP_KERNEL);
1da177e4 LT	62	if (!ei)
1da177e4 LT	63	return NULL;
13b41b09	64	ei->pid = NULL;
aed7a6c4	65	ei->fd = 0;
1da177e4 LT	66	ei->op.proc_get_link = NULL;
1da177e4 LT	67	ei->pde = NULL;
9043476f AV	68	ei->sysctl = NULL;
9043476f AV	69	ei->sysctl_entry = NULL;
3d3d35b1	70	ei->ns_ops = NULL;
1da177e4	71	inode = &ei->vfs_inode;
1da177e4 LT	72	return inode;
	73	}
	74
fa0d7e3d	75	static void proc_i_callback(struct rcu_head *head)
1da177e4	76	{
fa0d7e3d	77	struct inode *inode = container_of(head, struct inode, i_rcu);
1da177e4 LT	78	kmem_cache_free(proc_inode_cachep, PROC_I(inode));
	79	}
	80
fa0d7e3d NP	81	static void proc_destroy_inode(struct inode *inode)
	82	{
	83	call_rcu(&inode->i_rcu, proc_i_callback);
	84	}
	85
51cc5068	86	static void init_once(void *foo)
1da177e4 LT	87	{
	88	struct proc_inode ei = (struct proc_inode ) foo;
	89
a35afb83	90	inode_init_once(&ei->vfs_inode);
1da177e4	91	}
20c2df83	92
5bcd7ff9	93	void __init proc_init_inodecache(void)
1da177e4 LT	94	{
	95	proc_inode_cachep = kmem_cache_create("proc_inode_cache",
	96	sizeof(struct proc_inode),
fffb60f9	97	0, (SLAB_RECLAIM_ACCOUNT\|
5d097056 VD	98	SLAB_MEM_SPREAD\|SLAB_ACCOUNT\|
5d097056 VD	99	SLAB_PANIC),
20c2df83	100	init_once);
1da177e4 LT	101	}
1da177e4 LT	102
97412950 VK	103	static int proc_show_options(struct seq_file seq, struct dentry root)
97412950 VK	104	{
0499680a VK	105	struct super_block *sb = root->d_sb;
	106	struct pid_namespace *pid = sb->s_fs_info;
	107
dcb0f222 EB	108	if (!gid_eq(pid->pid_gid, GLOBAL_ROOT_GID))
dcb0f222 EB	109	seq_printf(seq, ",gid=%u", from_kgid_munged(&init_user_ns, pid->pid_gid));
796f571b	110	if (pid->hide_pid != HIDEPID_OFF)
0499680a VK	111	seq_printf(seq, ",hidepid=%u", pid->hide_pid);
0499680a VK	112
97412950 VK	113	return 0;
	114	}
	115
ee9b6d61	116	static const struct super_operations proc_sops = {
1da177e4 LT	117	.alloc_inode = proc_alloc_inode,
1da177e4 LT	118	.destroy_inode = proc_destroy_inode,
1da177e4	119	.drop_inode = generic_delete_inode,
8267952b	120	.evict_inode = proc_evict_inode,
1da177e4	121	.statfs = simple_statfs,
97412950 VK	122	.remount_fs = proc_remount,
97412950 VK	123	.show_options = proc_show_options,
1da177e4 LT	124	};
1da177e4 LT	125
866ad9a7 AV	126	enum {BIAS = -1U<<31};
	127
	128	static inline int use_pde(struct proc_dir_entry *pde)
	129	{
05c0ae21	130	return atomic_inc_unless_negative(&pde->in_use);
881adb85 AD	131	}
881adb85 AD	132
866ad9a7	133	static void unuse_pde(struct proc_dir_entry *pde)
881adb85	134	{
05c0ae21 AV	135	if (atomic_dec_return(&pde->in_use) == BIAS)
05c0ae21 AV	136	complete(pde->pde_unload_completion);
786d7e16 AD	137	}
786d7e16 AD	138
ca469f35 AV	139	/* pde is locked */
	140	static void close_pdeo(struct proc_dir_entry pde, struct pde_opener pdeo)
	141	{
492b2da6 AD	142	/*
	143	* close() (proc_reg_release()) can't delete an entry and proceed:
	144	* ->release hook needs to be available at the right moment.
	145	*
	146	* rmmod (remove_proc_entry() et al) can't delete an entry and proceed:
	147	* "struct file" needs to be available at the right moment.
	148	*
	149	* Therefore, first process to enter this function does ->release() and
	150	* signals its completion to the other process which does nothing.
	151	*/
05c0ae21	152	if (pdeo->closing) {
ca469f35	153	/* somebody else is doing that, just wait */
05c0ae21 AV	154	DECLARE_COMPLETION_ONSTACK(c);
05c0ae21 AV	155	pdeo->c = &c;
ca469f35	156	spin_unlock(&pde->pde_unload_lock);
05c0ae21	157	wait_for_completion(&c);
ca469f35	158	spin_lock(&pde->pde_unload_lock);
ca469f35 AV	159	} else {
ca469f35 AV	160	struct file *file;
f5887c71	161	pdeo->closing = true;
ca469f35 AV	162	spin_unlock(&pde->pde_unload_lock);
	163	file = pdeo->file;
	164	pde->proc_fops->release(file_inode(file), file);
	165	spin_lock(&pde->pde_unload_lock);
492b2da6	166	/* After ->release. */
06a0c417	167	list_del(&pdeo->lh);
05c0ae21 AV	168	if (pdeo->c)
05c0ae21 AV	169	complete(pdeo->c);
ca469f35	170	kfree(pdeo);
05c0ae21	171	}
ca469f35 AV	172	}
ca469f35 AV	173
866ad9a7	174	void proc_entry_rundown(struct proc_dir_entry *de)
786d7e16	175	{
05c0ae21	176	DECLARE_COMPLETION_ONSTACK(c);
866ad9a7	177	/* Wait until all existing callers into module are done. */
05c0ae21 AV	178	de->pde_unload_completion = &c;
	179	if (atomic_add_return(BIAS, &de->in_use) != BIAS)
	180	wait_for_completion(&c);
786d7e16	181
492b2da6 AD	182	/* ->pde_openers list can't grow from now on. */
492b2da6 AD	183
05c0ae21	184	spin_lock(&de->pde_unload_lock);
866ad9a7 AV	185	while (!list_empty(&de->pde_openers)) {
866ad9a7 AV	186	struct pde_opener *pdeo;
866ad9a7	187	pdeo = list_first_entry(&de->pde_openers, struct pde_opener, lh);
ca469f35	188	close_pdeo(de, pdeo);
866ad9a7 AV	189	}
866ad9a7 AV	190	spin_unlock(&de->pde_unload_lock);
786d7e16 AD	191	}
786d7e16 AD	192
866ad9a7 AV	193	static loff_t proc_reg_llseek(struct file *file, loff_t offset, int whence)
	194	{
	195	struct proc_dir_entry *pde = PDE(file_inode(file));
	196	loff_t rv = -EINVAL;
	197	if (use_pde(pde)) {
	198	loff_t (llseek)(struct file , loff_t, int);
	199	llseek = pde->proc_fops->llseek;
	200	if (!llseek)
	201	llseek = default_llseek;
	202	rv = llseek(file, offset, whence);
	203	unuse_pde(pde);
	204	}
786d7e16 AD	205	return rv;
	206	}
	207
866ad9a7	208	static ssize_t proc_reg_read(struct file file, char __user buf, size_t count, loff_t *ppos)
786d7e16	209	{
866ad9a7	210	ssize_t (read)(struct file , char __user , size_t, loff_t );
496ad9aa	211	struct proc_dir_entry *pde = PDE(file_inode(file));
786d7e16	212	ssize_t rv = -EIO;
866ad9a7 AV	213	if (use_pde(pde)) {
	214	read = pde->proc_fops->read;
	215	if (read)
	216	rv = read(file, buf, count, ppos);
	217	unuse_pde(pde);
786d7e16	218	}
866ad9a7 AV	219	return rv;
866ad9a7 AV	220	}
786d7e16	221
866ad9a7 AV	222	static ssize_t proc_reg_write(struct file file, const char __user buf, size_t count, loff_t *ppos)
	223	{
	224	ssize_t (write)(struct file , const char __user , size_t, loff_t );
	225	struct proc_dir_entry *pde = PDE(file_inode(file));
	226	ssize_t rv = -EIO;
	227	if (use_pde(pde)) {
	228	write = pde->proc_fops->write;
	229	if (write)
	230	rv = write(file, buf, count, ppos);
	231	unuse_pde(pde);
	232	}
786d7e16 AD	233	return rv;
	234	}
	235
	236	static unsigned int proc_reg_poll(struct file file, struct poll_table_struct pts)
	237	{
496ad9aa	238	struct proc_dir_entry *pde = PDE(file_inode(file));
dd23aae4	239	unsigned int rv = DEFAULT_POLLMASK;
786d7e16	240	unsigned int (poll)(struct file , struct poll_table_struct *);
866ad9a7 AV	241	if (use_pde(pde)) {
	242	poll = pde->proc_fops->poll;
	243	if (poll)
	244	rv = poll(file, pts);
	245	unuse_pde(pde);
786d7e16	246	}
786d7e16 AD	247	return rv;
	248	}
	249
	250	static long proc_reg_unlocked_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
	251	{
496ad9aa	252	struct proc_dir_entry *pde = PDE(file_inode(file));
786d7e16	253	long rv = -ENOTTY;
b19dd42f	254	long (ioctl)(struct file , unsigned int, unsigned long);
866ad9a7 AV	255	if (use_pde(pde)) {
	256	ioctl = pde->proc_fops->unlocked_ioctl;
	257	if (ioctl)
	258	rv = ioctl(file, cmd, arg);
	259	unuse_pde(pde);
786d7e16	260	}
786d7e16 AD	261	return rv;
	262	}
	263
	264	#ifdef CONFIG_COMPAT
	265	static long proc_reg_compat_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
	266	{
496ad9aa	267	struct proc_dir_entry *pde = PDE(file_inode(file));
786d7e16 AD	268	long rv = -ENOTTY;
786d7e16 AD	269	long (compat_ioctl)(struct file , unsigned int, unsigned long);
866ad9a7 AV	270	if (use_pde(pde)) {
	271	compat_ioctl = pde->proc_fops->compat_ioctl;
	272	if (compat_ioctl)
	273	rv = compat_ioctl(file, cmd, arg);
	274	unuse_pde(pde);
786d7e16	275	}
786d7e16 AD	276	return rv;
	277	}
	278	#endif
	279
	280	static int proc_reg_mmap(struct file file, struct vm_area_struct vma)
	281	{
496ad9aa	282	struct proc_dir_entry *pde = PDE(file_inode(file));
786d7e16 AD	283	int rv = -EIO;
786d7e16 AD	284	int (mmap)(struct file , struct vm_area_struct *);
866ad9a7 AV	285	if (use_pde(pde)) {
	286	mmap = pde->proc_fops->mmap;
	287	if (mmap)
	288	rv = mmap(file, vma);
	289	unuse_pde(pde);
786d7e16	290	}
786d7e16 AD	291	return rv;
	292	}
	293
5721cf84 HD	294	static unsigned long
	295	proc_reg_get_unmapped_area(struct file *file, unsigned long orig_addr,
	296	unsigned long len, unsigned long pgoff,
	297	unsigned long flags)
c4fe2448 AD	298	{
c4fe2448 AD	299	struct proc_dir_entry *pde = PDE(file_inode(file));
2cbe3b0a	300	unsigned long rv = -EIO;
ae5758a1	301
c4fe2448	302	if (use_pde(pde)) {
ae5758a1 JB	303	typeof(proc_reg_get_unmapped_area) *get_area;
	304
	305	get_area = pde->proc_fops->get_unmapped_area;
fad1a86e	306	#ifdef CONFIG_MMU
ae5758a1 JB	307	if (!get_area)
ae5758a1 JB	308	get_area = current->mm->get_unmapped_area;
fad1a86e	309	#endif
ae5758a1	310
5721cf84 HD	311	if (get_area)
5721cf84 HD	312	rv = get_area(file, orig_addr, len, pgoff, flags);
ae5758a1 JB	313	else
ae5758a1 JB	314	rv = orig_addr;
c4fe2448 AD	315	unuse_pde(pde);
	316	}
	317	return rv;
	318	}
	319
786d7e16 AD	320	static int proc_reg_open(struct inode inode, struct file file)
	321	{
	322	struct proc_dir_entry *pde = PDE(inode);
	323	int rv = 0;
	324	int (open)(struct inode , struct file *);
881adb85 AD	325	int (release)(struct inode , struct file *);
	326	struct pde_opener *pdeo;
	327
	328	/*
492b2da6 AD	329	* Ensure that
	330	* 1) PDE's ->release hook will be called no matter what
	331	* either normally by close()/->release, or forcefully by
	332	* rmmod/remove_proc_entry.
	333	*
	334	* 2) rmmod isn't blocked by opening file in /proc and sitting on
	335	* the descriptor (including "rmmod foo </proc/foo" scenario).
881adb85	336	*
492b2da6	337	* Save every "struct file" with custom ->release hook.
881adb85	338	*/
39a10ac2	339	pdeo = kmalloc(sizeof(struct pde_opener), GFP_KERNEL);
881adb85 AD	340	if (!pdeo)
881adb85 AD	341	return -ENOMEM;
786d7e16	342
866ad9a7	343	if (!use_pde(pde)) {
881adb85	344	kfree(pdeo);
d2857e79	345	return -ENOENT;
786d7e16	346	}
786d7e16	347	open = pde->proc_fops->open;
881adb85	348	release = pde->proc_fops->release;
786d7e16 AD	349
	350	if (open)
	351	rv = open(inode, file);
	352
881adb85 AD	353	if (rv == 0 && release) {
881adb85 AD	354	/* To know what to release. */
881adb85	355	pdeo->file = file;
39a10ac2 AD	356	pdeo->closing = false;
39a10ac2 AD	357	pdeo->c = NULL;
05c0ae21	358	spin_lock(&pde->pde_unload_lock);
881adb85	359	list_add(&pdeo->lh, &pde->pde_openers);
05c0ae21	360	spin_unlock(&pde->pde_unload_lock);
881adb85 AD	361	} else
881adb85 AD	362	kfree(pdeo);
05c0ae21 AV	363
05c0ae21 AV	364	unuse_pde(pde);
786d7e16 AD	365	return rv;
	366	}
	367
	368	static int proc_reg_release(struct inode inode, struct file file)
	369	{
	370	struct proc_dir_entry *pde = PDE(inode);
881adb85	371	struct pde_opener *pdeo;
786d7e16	372	spin_lock(&pde->pde_unload_lock);
ca469f35 AV	373	list_for_each_entry(pdeo, &pde->pde_openers, lh) {
	374	if (pdeo->file == file) {
	375	close_pdeo(pde, pdeo);
	376	break;
	377	}
881adb85	378	}
786d7e16	379	spin_unlock(&pde->pde_unload_lock);
ca469f35	380	return 0;
786d7e16 AD	381	}
	382
	383	static const struct file_operations proc_reg_file_ops = {
	384	.llseek = proc_reg_llseek,
	385	.read = proc_reg_read,
	386	.write = proc_reg_write,
	387	.poll = proc_reg_poll,
	388	.unlocked_ioctl = proc_reg_unlocked_ioctl,
	389	#ifdef CONFIG_COMPAT
	390	.compat_ioctl = proc_reg_compat_ioctl,
	391	#endif
	392	.mmap = proc_reg_mmap,
c4fe2448	393	.get_unmapped_area = proc_reg_get_unmapped_area,
786d7e16 AD	394	.open = proc_reg_open,
	395	.release = proc_reg_release,
	396	};
	397
778f3dd5 DM	398	#ifdef CONFIG_COMPAT
	399	static const struct file_operations proc_reg_file_ops_no_compat = {
	400	.llseek = proc_reg_llseek,
	401	.read = proc_reg_read,
	402	.write = proc_reg_write,
	403	.poll = proc_reg_poll,
	404	.unlocked_ioctl = proc_reg_unlocked_ioctl,
	405	.mmap = proc_reg_mmap,
c4fe2448	406	.get_unmapped_area = proc_reg_get_unmapped_area,
778f3dd5 DM	407	.open = proc_reg_open,
	408	.release = proc_reg_release,
	409	};
	410	#endif
	411
fceef393 AV	412	static void proc_put_link(void *p)
	413	{
	414	unuse_pde(p);
	415	}
	416
6b255391	417	static const char proc_get_link(struct dentry dentry,
fceef393 AV	418	struct inode *inode,
fceef393 AV	419	struct delayed_call *done)
7e0e953b	420	{
6b255391	421	struct proc_dir_entry *pde = PDE(inode);
7e0e953b AV	422	if (unlikely(!use_pde(pde)))
7e0e953b AV	423	return ERR_PTR(-EINVAL);
fceef393	424	set_delayed_call(done, proc_put_link, pde);
680baacb	425	return pde->data;
7e0e953b AV	426	}
7e0e953b AV	427
7e0e953b	428	const struct inode_operations proc_link_inode_operations = {
6b255391	429	.get_link = proc_get_link,
7e0e953b AV	430	};
7e0e953b AV	431
6d1b6e4e	432	struct inode proc_get_inode(struct super_block sb, struct proc_dir_entry *de)
1da177e4	433	{
51f0885e	434	struct inode *inode = new_inode_pseudo(sb);
1da177e4	435
51f0885e LT	436	if (inode) {
51f0885e LT	437	inode->i_ino = de->low_ino;
078cd827	438	inode->i_mtime = inode->i_atime = inode->i_ctime = current_time(inode);
a1d4aebb	439	PROC_I(inode)->pde = de;
5e971dce	440
eb6d38d5 EB	441	if (is_empty_pde(de)) {
	442	make_empty_dir_inode(inode);
	443	return inode;
	444	}
5e971dce AD	445	if (de->mode) {
	446	inode->i_mode = de->mode;
	447	inode->i_uid = de->uid;
	448	inode->i_gid = de->gid;
	449	}
	450	if (de->size)
	451	inode->i_size = de->size;
	452	if (de->nlink)
bfe86848	453	set_nlink(inode, de->nlink);
b6cdc731 AV	454	WARN_ON(!de->proc_iops);
b6cdc731 AV	455	inode->i_op = de->proc_iops;
5e971dce AD	456	if (de->proc_fops) {
5e971dce AD	457	if (S_ISREG(inode->i_mode)) {
778f3dd5	458	#ifdef CONFIG_COMPAT
5e971dce AD	459	if (!de->proc_fops->compat_ioctl)
	460	inode->i_fop =
	461	&proc_reg_file_ops_no_compat;
	462	else
778f3dd5	463	#endif
5e971dce AD	464	inode->i_fop = &proc_reg_file_ops;
	465	} else {
	466	inode->i_fop = de->proc_fops;
778f3dd5	467	}
786d7e16	468	}
99b76233	469	} else
135d5655	470	pde_put(de);
1da177e4	471	return inode;
d3d009cb	472	}
1da177e4	473
e94591d0	474	int proc_fill_super(struct super_block s, void data, int silent)
1da177e4	475	{
e94591d0	476	struct pid_namespace *ns = get_pid_ns(s->s_fs_info);
87e0aab3	477	struct inode *root_inode;
0097875b	478	int ret;
87e0aab3	479
e94591d0 EB	480	if (!proc_parse_options(data, ns))
	481	return -EINVAL;
	482
a2982cc9 EB	483	/* User space would break if executables or devices appear on proc */
a2982cc9 EB	484	s->s_iflags \|= SB_I_USERNS_VISIBLE \| SB_I_NOEXEC \| SB_I_NODEV;
92d03285	485	s->s_flags \|= MS_NODIRATIME \| MS_NOSUID \| MS_NOEXEC;
1da177e4 LT	486	s->s_blocksize = 1024;
	487	s->s_blocksize_bits = 10;
	488	s->s_magic = PROC_SUPER_MAGIC;
	489	s->s_op = &proc_sops;
	490	s->s_time_gran = 1;
a867d734 LT	491
	492	/*
	493	* procfs isn't actually a stacking filesystem; however, there is
	494	* too much magic going on inside it to permit stacking things on
	495	* top of it
	496	*/
	497	s->s_stack_depth = FILESYSTEM_MAX_STACK_DEPTH;
1da177e4	498
135d5655	499	pde_get(&proc_root);
87e0aab3 MP	500	root_inode = proc_get_inode(s, &proc_root);
87e0aab3 MP	501	if (!root_inode) {
87ebdc00	502	pr_err("proc_fill_super: get root inode failed\n");
87e0aab3 MP	503	return -ENOMEM;
87e0aab3 MP	504	}
1da177e4	505
87e0aab3 MP	506	s->s_root = d_make_root(root_inode);
87e0aab3 MP	507	if (!s->s_root) {
87ebdc00	508	pr_err("proc_fill_super: allocate dentry failed\n");
87e0aab3 MP	509	return -ENOMEM;
	510	}
	511
0097875b EB	512	ret = proc_setup_self(s);
	513	if (ret) {
	514	return ret;
	515	}
	516	return proc_setup_thread_self(s);
1da177e4	517	}