[linux-2.6-block.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 <asm/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);
		sysctl_head_put(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 = (struct proc_inode *)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;
	inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME;
	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 != 0)
		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)
{
	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 = 1;
		spin_unlock(&pde->pde_unload_lock);
		file = pdeo->file;
		pde->proc_fops->release(file_inode(file), file);
		spin_lock(&pde->pde_unload_lock);
		list_del_init(&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);

	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;

	/*
	 * What for, you ask? Well, we can have open, rmmod, remove_proc_entry
	 * sequence. ->release won't be called because ->proc_fops will be
	 * cleared. Depending on complexity of ->release, consequences vary.
	 *
	 * We can't wait for mercy when close will be done for real, it's
	 * deadlockable: rmmod foo </proc/foo . So, we're going to do ->release
	 * by hand in remove_proc_entry(). For this, save opener's credentials
	 * for later.
	 */
	pdeo = kzalloc(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;
		/* Strictly for "too late" ->release in proc_reg_release(). */
		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 = {
	.readlink	= generic_readlink,
	.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;
		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
1da177e4 LT	27	#include <asm/uaccess.h>
1da177e4 LT	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);
dfef6dcd AV	46	head = PROC_I(inode)->sysctl;
dfef6dcd AV	47	if (head) {
1c44dbc8	48	RCU_INIT_POINTER(PROC_I(inode)->sysctl, NULL);
dfef6dcd AV	49	sysctl_head_put(head);
dfef6dcd AV	50	}
1da177e4 LT	51	}
1da177e4 LT	52
e18b890b	53	static struct kmem_cache * proc_inode_cachep;
1da177e4 LT	54
	55	static struct inode proc_alloc_inode(struct super_block sb)
	56	{
	57	struct proc_inode *ei;
	58	struct inode *inode;
	59
e94b1766	60	ei = (struct proc_inode *)kmem_cache_alloc(proc_inode_cachep, GFP_KERNEL);
1da177e4 LT	61	if (!ei)
1da177e4 LT	62	return NULL;
13b41b09	63	ei->pid = NULL;
aed7a6c4	64	ei->fd = 0;
1da177e4 LT	65	ei->op.proc_get_link = NULL;
1da177e4 LT	66	ei->pde = NULL;
9043476f AV	67	ei->sysctl = NULL;
9043476f AV	68	ei->sysctl_entry = NULL;
3d3d35b1	69	ei->ns_ops = NULL;
1da177e4 LT	70	inode = &ei->vfs_inode;
	71	inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME;
	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));
0499680a VK	110	if (pid->hide_pid != 0)
	111	seq_printf(seq, ",hidepid=%u", pid->hide_pid);
	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	{
05c0ae21	142	if (pdeo->closing) {
ca469f35	143	/* somebody else is doing that, just wait */
05c0ae21 AV	144	DECLARE_COMPLETION_ONSTACK(c);
05c0ae21 AV	145	pdeo->c = &c;
ca469f35	146	spin_unlock(&pde->pde_unload_lock);
05c0ae21	147	wait_for_completion(&c);
ca469f35	148	spin_lock(&pde->pde_unload_lock);
ca469f35 AV	149	} else {
ca469f35 AV	150	struct file *file;
05c0ae21	151	pdeo->closing = 1;
ca469f35 AV	152	spin_unlock(&pde->pde_unload_lock);
	153	file = pdeo->file;
	154	pde->proc_fops->release(file_inode(file), file);
	155	spin_lock(&pde->pde_unload_lock);
	156	list_del_init(&pdeo->lh);
05c0ae21 AV	157	if (pdeo->c)
05c0ae21 AV	158	complete(pdeo->c);
ca469f35	159	kfree(pdeo);
05c0ae21	160	}
ca469f35 AV	161	}
ca469f35 AV	162
866ad9a7	163	void proc_entry_rundown(struct proc_dir_entry *de)
786d7e16	164	{
05c0ae21	165	DECLARE_COMPLETION_ONSTACK(c);
866ad9a7	166	/* Wait until all existing callers into module are done. */
05c0ae21 AV	167	de->pde_unload_completion = &c;
	168	if (atomic_add_return(BIAS, &de->in_use) != BIAS)
	169	wait_for_completion(&c);
786d7e16	170
05c0ae21	171	spin_lock(&de->pde_unload_lock);
866ad9a7 AV	172	while (!list_empty(&de->pde_openers)) {
866ad9a7 AV	173	struct pde_opener *pdeo;
866ad9a7	174	pdeo = list_first_entry(&de->pde_openers, struct pde_opener, lh);
ca469f35	175	close_pdeo(de, pdeo);
866ad9a7 AV	176	}
866ad9a7 AV	177	spin_unlock(&de->pde_unload_lock);
786d7e16 AD	178	}
786d7e16 AD	179
866ad9a7 AV	180	static loff_t proc_reg_llseek(struct file *file, loff_t offset, int whence)
	181	{
	182	struct proc_dir_entry *pde = PDE(file_inode(file));
	183	loff_t rv = -EINVAL;
	184	if (use_pde(pde)) {
	185	loff_t (llseek)(struct file , loff_t, int);
	186	llseek = pde->proc_fops->llseek;
	187	if (!llseek)
	188	llseek = default_llseek;
	189	rv = llseek(file, offset, whence);
	190	unuse_pde(pde);
	191	}
786d7e16 AD	192	return rv;
	193	}
	194
866ad9a7	195	static ssize_t proc_reg_read(struct file file, char __user buf, size_t count, loff_t *ppos)
786d7e16	196	{
866ad9a7	197	ssize_t (read)(struct file , char __user , size_t, loff_t );
496ad9aa	198	struct proc_dir_entry *pde = PDE(file_inode(file));
786d7e16	199	ssize_t rv = -EIO;
866ad9a7 AV	200	if (use_pde(pde)) {
	201	read = pde->proc_fops->read;
	202	if (read)
	203	rv = read(file, buf, count, ppos);
	204	unuse_pde(pde);
786d7e16	205	}
866ad9a7 AV	206	return rv;
866ad9a7 AV	207	}
786d7e16	208
866ad9a7 AV	209	static ssize_t proc_reg_write(struct file file, const char __user buf, size_t count, loff_t *ppos)
	210	{
	211	ssize_t (write)(struct file , const char __user , size_t, loff_t );
	212	struct proc_dir_entry *pde = PDE(file_inode(file));
	213	ssize_t rv = -EIO;
	214	if (use_pde(pde)) {
	215	write = pde->proc_fops->write;
	216	if (write)
	217	rv = write(file, buf, count, ppos);
	218	unuse_pde(pde);
	219	}
786d7e16 AD	220	return rv;
	221	}
	222
	223	static unsigned int proc_reg_poll(struct file file, struct poll_table_struct pts)
	224	{
496ad9aa	225	struct proc_dir_entry *pde = PDE(file_inode(file));
dd23aae4	226	unsigned int rv = DEFAULT_POLLMASK;
786d7e16	227	unsigned int (poll)(struct file , struct poll_table_struct *);
866ad9a7 AV	228	if (use_pde(pde)) {
	229	poll = pde->proc_fops->poll;
	230	if (poll)
	231	rv = poll(file, pts);
	232	unuse_pde(pde);
786d7e16	233	}
786d7e16 AD	234	return rv;
	235	}
	236
	237	static long proc_reg_unlocked_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
	238	{
496ad9aa	239	struct proc_dir_entry *pde = PDE(file_inode(file));
786d7e16	240	long rv = -ENOTTY;
b19dd42f	241	long (ioctl)(struct file , unsigned int, unsigned long);
866ad9a7 AV	242	if (use_pde(pde)) {
	243	ioctl = pde->proc_fops->unlocked_ioctl;
	244	if (ioctl)
	245	rv = ioctl(file, cmd, arg);
	246	unuse_pde(pde);
786d7e16	247	}
786d7e16 AD	248	return rv;
	249	}
	250
	251	#ifdef CONFIG_COMPAT
	252	static long proc_reg_compat_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
	253	{
496ad9aa	254	struct proc_dir_entry *pde = PDE(file_inode(file));
786d7e16 AD	255	long rv = -ENOTTY;
786d7e16 AD	256	long (compat_ioctl)(struct file , unsigned int, unsigned long);
866ad9a7 AV	257	if (use_pde(pde)) {
	258	compat_ioctl = pde->proc_fops->compat_ioctl;
	259	if (compat_ioctl)
	260	rv = compat_ioctl(file, cmd, arg);
	261	unuse_pde(pde);
786d7e16	262	}
786d7e16 AD	263	return rv;
	264	}
	265	#endif
	266
	267	static int proc_reg_mmap(struct file file, struct vm_area_struct vma)
	268	{
496ad9aa	269	struct proc_dir_entry *pde = PDE(file_inode(file));
786d7e16 AD	270	int rv = -EIO;
786d7e16 AD	271	int (mmap)(struct file , struct vm_area_struct *);
866ad9a7 AV	272	if (use_pde(pde)) {
	273	mmap = pde->proc_fops->mmap;
	274	if (mmap)
	275	rv = mmap(file, vma);
	276	unuse_pde(pde);
786d7e16	277	}
786d7e16 AD	278	return rv;
	279	}
	280
5721cf84 HD	281	static unsigned long
	282	proc_reg_get_unmapped_area(struct file *file, unsigned long orig_addr,
	283	unsigned long len, unsigned long pgoff,
	284	unsigned long flags)
c4fe2448 AD	285	{
c4fe2448 AD	286	struct proc_dir_entry *pde = PDE(file_inode(file));
2cbe3b0a	287	unsigned long rv = -EIO;
ae5758a1	288
c4fe2448	289	if (use_pde(pde)) {
ae5758a1 JB	290	typeof(proc_reg_get_unmapped_area) *get_area;
	291
	292	get_area = pde->proc_fops->get_unmapped_area;
fad1a86e	293	#ifdef CONFIG_MMU
ae5758a1 JB	294	if (!get_area)
ae5758a1 JB	295	get_area = current->mm->get_unmapped_area;
fad1a86e	296	#endif
ae5758a1	297
5721cf84 HD	298	if (get_area)
5721cf84 HD	299	rv = get_area(file, orig_addr, len, pgoff, flags);
ae5758a1 JB	300	else
ae5758a1 JB	301	rv = orig_addr;
c4fe2448 AD	302	unuse_pde(pde);
	303	}
	304	return rv;
	305	}
	306
786d7e16 AD	307	static int proc_reg_open(struct inode inode, struct file file)
	308	{
	309	struct proc_dir_entry *pde = PDE(inode);
	310	int rv = 0;
	311	int (open)(struct inode , struct file *);
881adb85 AD	312	int (release)(struct inode , struct file *);
	313	struct pde_opener *pdeo;
	314
	315	/*
	316	* What for, you ask? Well, we can have open, rmmod, remove_proc_entry
	317	* sequence. ->release won't be called because ->proc_fops will be
	318	* cleared. Depending on complexity of ->release, consequences vary.
	319	*
	320	* We can't wait for mercy when close will be done for real, it's
	321	* deadlockable: rmmod foo </proc/foo . So, we're going to do ->release
	322	* by hand in remove_proc_entry(). For this, save opener's credentials
	323	* for later.
	324	*/
05c0ae21	325	pdeo = kzalloc(sizeof(struct pde_opener), GFP_KERNEL);
881adb85 AD	326	if (!pdeo)
881adb85 AD	327	return -ENOMEM;
786d7e16	328
866ad9a7	329	if (!use_pde(pde)) {
881adb85	330	kfree(pdeo);
d2857e79	331	return -ENOENT;
786d7e16	332	}
786d7e16	333	open = pde->proc_fops->open;
881adb85	334	release = pde->proc_fops->release;
786d7e16 AD	335
	336	if (open)
	337	rv = open(inode, file);
	338
881adb85 AD	339	if (rv == 0 && release) {
881adb85 AD	340	/* To know what to release. */
881adb85 AD	341	pdeo->file = file;
881adb85 AD	342	/* Strictly for "too late" ->release in proc_reg_release(). */
05c0ae21	343	spin_lock(&pde->pde_unload_lock);
881adb85	344	list_add(&pdeo->lh, &pde->pde_openers);
05c0ae21	345	spin_unlock(&pde->pde_unload_lock);
881adb85 AD	346	} else
881adb85 AD	347	kfree(pdeo);
05c0ae21 AV	348
05c0ae21 AV	349	unuse_pde(pde);
786d7e16 AD	350	return rv;
	351	}
	352
	353	static int proc_reg_release(struct inode inode, struct file file)
	354	{
	355	struct proc_dir_entry *pde = PDE(inode);
881adb85	356	struct pde_opener *pdeo;
786d7e16	357	spin_lock(&pde->pde_unload_lock);
ca469f35 AV	358	list_for_each_entry(pdeo, &pde->pde_openers, lh) {
	359	if (pdeo->file == file) {
	360	close_pdeo(pde, pdeo);
	361	break;
	362	}
881adb85	363	}
786d7e16	364	spin_unlock(&pde->pde_unload_lock);
ca469f35	365	return 0;
786d7e16 AD	366	}
	367
	368	static const struct file_operations proc_reg_file_ops = {
	369	.llseek = proc_reg_llseek,
	370	.read = proc_reg_read,
	371	.write = proc_reg_write,
	372	.poll = proc_reg_poll,
	373	.unlocked_ioctl = proc_reg_unlocked_ioctl,
	374	#ifdef CONFIG_COMPAT
	375	.compat_ioctl = proc_reg_compat_ioctl,
	376	#endif
	377	.mmap = proc_reg_mmap,
c4fe2448	378	.get_unmapped_area = proc_reg_get_unmapped_area,
786d7e16 AD	379	.open = proc_reg_open,
	380	.release = proc_reg_release,
	381	};
	382
778f3dd5 DM	383	#ifdef CONFIG_COMPAT
	384	static const struct file_operations proc_reg_file_ops_no_compat = {
	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	.mmap = proc_reg_mmap,
c4fe2448	391	.get_unmapped_area = proc_reg_get_unmapped_area,
778f3dd5 DM	392	.open = proc_reg_open,
	393	.release = proc_reg_release,
	394	};
	395	#endif
	396
fceef393 AV	397	static void proc_put_link(void *p)
	398	{
	399	unuse_pde(p);
	400	}
	401
6b255391	402	static const char proc_get_link(struct dentry dentry,
fceef393 AV	403	struct inode *inode,
fceef393 AV	404	struct delayed_call *done)
7e0e953b	405	{
6b255391	406	struct proc_dir_entry *pde = PDE(inode);
7e0e953b AV	407	if (unlikely(!use_pde(pde)))
7e0e953b AV	408	return ERR_PTR(-EINVAL);
fceef393	409	set_delayed_call(done, proc_put_link, pde);
680baacb	410	return pde->data;
7e0e953b AV	411	}
7e0e953b AV	412
7e0e953b AV	413	const struct inode_operations proc_link_inode_operations = {
7e0e953b AV	414	.readlink = generic_readlink,
6b255391	415	.get_link = proc_get_link,
7e0e953b AV	416	};
7e0e953b AV	417
6d1b6e4e	418	struct inode proc_get_inode(struct super_block sb, struct proc_dir_entry *de)
1da177e4	419	{
51f0885e	420	struct inode *inode = new_inode_pseudo(sb);
1da177e4	421
51f0885e LT	422	if (inode) {
51f0885e LT	423	inode->i_ino = de->low_ino;
a1d4aebb	424	inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME;
a1d4aebb	425	PROC_I(inode)->pde = de;
5e971dce	426
eb6d38d5 EB	427	if (is_empty_pde(de)) {
	428	make_empty_dir_inode(inode);
	429	return inode;
	430	}
5e971dce AD	431	if (de->mode) {
	432	inode->i_mode = de->mode;
	433	inode->i_uid = de->uid;
	434	inode->i_gid = de->gid;
	435	}
	436	if (de->size)
	437	inode->i_size = de->size;
	438	if (de->nlink)
bfe86848	439	set_nlink(inode, de->nlink);
b6cdc731 AV	440	WARN_ON(!de->proc_iops);
b6cdc731 AV	441	inode->i_op = de->proc_iops;
5e971dce AD	442	if (de->proc_fops) {
5e971dce AD	443	if (S_ISREG(inode->i_mode)) {
778f3dd5	444	#ifdef CONFIG_COMPAT
5e971dce AD	445	if (!de->proc_fops->compat_ioctl)
	446	inode->i_fop =
	447	&proc_reg_file_ops_no_compat;
	448	else
778f3dd5	449	#endif
5e971dce AD	450	inode->i_fop = &proc_reg_file_ops;
	451	} else {
	452	inode->i_fop = de->proc_fops;
778f3dd5	453	}
786d7e16	454	}
99b76233	455	} else
135d5655	456	pde_put(de);
1da177e4	457	return inode;
d3d009cb	458	}
1da177e4	459
e94591d0	460	int proc_fill_super(struct super_block s, void data, int silent)
1da177e4	461	{
e94591d0	462	struct pid_namespace *ns = get_pid_ns(s->s_fs_info);
87e0aab3	463	struct inode *root_inode;
0097875b	464	int ret;
87e0aab3	465
e94591d0 EB	466	if (!proc_parse_options(data, ns))
	467	return -EINVAL;
	468
a2982cc9 EB	469	/* User space would break if executables or devices appear on proc */
a2982cc9 EB	470	s->s_iflags \|= SB_I_USERNS_VISIBLE \| SB_I_NOEXEC \| SB_I_NODEV;
92d03285	471	s->s_flags \|= MS_NODIRATIME \| MS_NOSUID \| MS_NOEXEC;
1da177e4 LT	472	s->s_blocksize = 1024;
	473	s->s_blocksize_bits = 10;
	474	s->s_magic = PROC_SUPER_MAGIC;
	475	s->s_op = &proc_sops;
	476	s->s_time_gran = 1;
a867d734 LT	477
	478	/*
	479	* procfs isn't actually a stacking filesystem; however, there is
	480	* too much magic going on inside it to permit stacking things on
	481	* top of it
	482	*/
	483	s->s_stack_depth = FILESYSTEM_MAX_STACK_DEPTH;
1da177e4	484
135d5655	485	pde_get(&proc_root);
87e0aab3 MP	486	root_inode = proc_get_inode(s, &proc_root);
87e0aab3 MP	487	if (!root_inode) {
87ebdc00	488	pr_err("proc_fill_super: get root inode failed\n");
87e0aab3 MP	489	return -ENOMEM;
87e0aab3 MP	490	}
1da177e4	491
87e0aab3 MP	492	s->s_root = d_make_root(root_inode);
87e0aab3 MP	493	if (!s->s_root) {
87ebdc00	494	pr_err("proc_fill_super: allocate dentry failed\n");
87e0aab3 MP	495	return -ENOMEM;
	496	}
	497
0097875b EB	498	ret = proc_setup_self(s);
	499	if (ret) {
	500	return ret;
	501	}
	502	return proc_setup_thread_self(s);
1da177e4	503	}