[linux.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;
	const struct proc_ns_operations *ns_ops;
	void *ns;

	truncate_inode_pages(&inode->i_data, 0);
	clear_inode(inode);

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

	/* Let go of any associated proc directory entry */
	de = PROC_I(inode)->pde;
	if (de)
		pde_put(de);
	head = PROC_I(inode)->sysctl;
	if (head) {
		rcu_assign_pointer(PROC_I(inode)->sysctl, NULL);
		sysctl_head_put(head);
	}
	/* Release any associated namespace */
	ns_ops = PROC_I(inode)->ns.ns_ops;
	ns = PROC_I(inode)->ns.ns;
	if (ns_ops && ns)
		ns_ops->put(ns);
}

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.ns = NULL;
	ei->ns.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_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 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,
	.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,
	.open		= proc_reg_open,
	.release	= proc_reg_release,
};
#endif

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 (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)
{
	struct inode *root_inode;

	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;
	
	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;
	}

	return proc_setup_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;
6b4e306a	35	const struct proc_ns_operations *ns_ops;
bf056bfa	36	void *ns;
1da177e4	37
fef26658	38	truncate_inode_pages(&inode->i_data, 0);
dbd5768f	39	clear_inode(inode);
fef26658	40
99f89551	41	/* Stop tracking associated processes */
13b41b09	42	put_pid(PROC_I(inode)->pid);
1da177e4 LT	43
	44	/* Let go of any associated proc directory entry */
	45	de = PROC_I(inode)->pde;
99b76233	46	if (de)
135d5655	47	pde_put(de);
dfef6dcd AV	48	head = PROC_I(inode)->sysctl;
	49	if (head) {
	50	rcu_assign_pointer(PROC_I(inode)->sysctl, NULL);
	51	sysctl_head_put(head);
	52	}
6b4e306a	53	/* Release any associated namespace */
0bb80f24 DH	54	ns_ops = PROC_I(inode)->ns.ns_ops;
0bb80f24 DH	55	ns = PROC_I(inode)->ns.ns;
bf056bfa EB	56	if (ns_ops && ns)
bf056bfa EB	57	ns_ops->put(ns);
1da177e4 LT	58	}
1da177e4 LT	59
e18b890b	60	static struct kmem_cache * proc_inode_cachep;
1da177e4 LT	61
	62	static struct inode proc_alloc_inode(struct super_block sb)
	63	{
	64	struct proc_inode *ei;
	65	struct inode *inode;
	66
e94b1766	67	ei = (struct proc_inode *)kmem_cache_alloc(proc_inode_cachep, GFP_KERNEL);
1da177e4 LT	68	if (!ei)
1da177e4 LT	69	return NULL;
13b41b09	70	ei->pid = NULL;
aed7a6c4	71	ei->fd = 0;
1da177e4 LT	72	ei->op.proc_get_link = NULL;
1da177e4 LT	73	ei->pde = NULL;
9043476f AV	74	ei->sysctl = NULL;
9043476f AV	75	ei->sysctl_entry = NULL;
0bb80f24 DH	76	ei->ns.ns = NULL;
0bb80f24 DH	77	ei->ns.ns_ops = NULL;
1da177e4 LT	78	inode = &ei->vfs_inode;
	79	inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME;
	80	return inode;
	81	}
	82
fa0d7e3d	83	static void proc_i_callback(struct rcu_head *head)
1da177e4	84	{
fa0d7e3d	85	struct inode *inode = container_of(head, struct inode, i_rcu);
1da177e4 LT	86	kmem_cache_free(proc_inode_cachep, PROC_I(inode));
	87	}
	88
fa0d7e3d NP	89	static void proc_destroy_inode(struct inode *inode)
	90	{
	91	call_rcu(&inode->i_rcu, proc_i_callback);
	92	}
	93
51cc5068	94	static void init_once(void *foo)
1da177e4 LT	95	{
	96	struct proc_inode ei = (struct proc_inode ) foo;
	97
a35afb83	98	inode_init_once(&ei->vfs_inode);
1da177e4	99	}
20c2df83	100
5bcd7ff9	101	void __init proc_init_inodecache(void)
1da177e4 LT	102	{
	103	proc_inode_cachep = kmem_cache_create("proc_inode_cache",
	104	sizeof(struct proc_inode),
fffb60f9	105	0, (SLAB_RECLAIM_ACCOUNT\|
040b5c6f	106	SLAB_MEM_SPREAD\|SLAB_PANIC),
20c2df83	107	init_once);
1da177e4 LT	108	}
1da177e4 LT	109
97412950 VK	110	static int proc_show_options(struct seq_file seq, struct dentry root)
97412950 VK	111	{
0499680a VK	112	struct super_block *sb = root->d_sb;
	113	struct pid_namespace *pid = sb->s_fs_info;
	114
dcb0f222 EB	115	if (!gid_eq(pid->pid_gid, GLOBAL_ROOT_GID))
dcb0f222 EB	116	seq_printf(seq, ",gid=%u", from_kgid_munged(&init_user_ns, pid->pid_gid));
0499680a VK	117	if (pid->hide_pid != 0)
	118	seq_printf(seq, ",hidepid=%u", pid->hide_pid);
	119
97412950 VK	120	return 0;
	121	}
	122
ee9b6d61	123	static const struct super_operations proc_sops = {
1da177e4 LT	124	.alloc_inode = proc_alloc_inode,
1da177e4 LT	125	.destroy_inode = proc_destroy_inode,
1da177e4	126	.drop_inode = generic_delete_inode,
8267952b	127	.evict_inode = proc_evict_inode,
1da177e4	128	.statfs = simple_statfs,
97412950 VK	129	.remount_fs = proc_remount,
97412950 VK	130	.show_options = proc_show_options,
1da177e4 LT	131	};
1da177e4 LT	132
866ad9a7 AV	133	enum {BIAS = -1U<<31};
	134
	135	static inline int use_pde(struct proc_dir_entry *pde)
	136	{
05c0ae21	137	return atomic_inc_unless_negative(&pde->in_use);
881adb85 AD	138	}
881adb85 AD	139
866ad9a7	140	static void unuse_pde(struct proc_dir_entry *pde)
881adb85	141	{
05c0ae21 AV	142	if (atomic_dec_return(&pde->in_use) == BIAS)
05c0ae21 AV	143	complete(pde->pde_unload_completion);
786d7e16 AD	144	}
786d7e16 AD	145
ca469f35 AV	146	/* pde is locked */
	147	static void close_pdeo(struct proc_dir_entry pde, struct pde_opener pdeo)
	148	{
05c0ae21	149	if (pdeo->closing) {
ca469f35	150	/* somebody else is doing that, just wait */
05c0ae21 AV	151	DECLARE_COMPLETION_ONSTACK(c);
05c0ae21 AV	152	pdeo->c = &c;
ca469f35	153	spin_unlock(&pde->pde_unload_lock);
05c0ae21	154	wait_for_completion(&c);
ca469f35	155	spin_lock(&pde->pde_unload_lock);
ca469f35 AV	156	} else {
ca469f35 AV	157	struct file *file;
05c0ae21	158	pdeo->closing = 1;
ca469f35 AV	159	spin_unlock(&pde->pde_unload_lock);
	160	file = pdeo->file;
	161	pde->proc_fops->release(file_inode(file), file);
	162	spin_lock(&pde->pde_unload_lock);
	163	list_del_init(&pdeo->lh);
05c0ae21 AV	164	if (pdeo->c)
05c0ae21 AV	165	complete(pdeo->c);
ca469f35	166	kfree(pdeo);
05c0ae21	167	}
ca469f35 AV	168	}
ca469f35 AV	169
866ad9a7	170	void proc_entry_rundown(struct proc_dir_entry *de)
786d7e16	171	{
05c0ae21	172	DECLARE_COMPLETION_ONSTACK(c);
866ad9a7	173	/* Wait until all existing callers into module are done. */
05c0ae21 AV	174	de->pde_unload_completion = &c;
	175	if (atomic_add_return(BIAS, &de->in_use) != BIAS)
	176	wait_for_completion(&c);
786d7e16	177
05c0ae21	178	spin_lock(&de->pde_unload_lock);
866ad9a7 AV	179	while (!list_empty(&de->pde_openers)) {
866ad9a7 AV	180	struct pde_opener *pdeo;
866ad9a7	181	pdeo = list_first_entry(&de->pde_openers, struct pde_opener, lh);
ca469f35	182	close_pdeo(de, pdeo);
866ad9a7 AV	183	}
866ad9a7 AV	184	spin_unlock(&de->pde_unload_lock);
786d7e16 AD	185	}
786d7e16 AD	186
866ad9a7 AV	187	static loff_t proc_reg_llseek(struct file *file, loff_t offset, int whence)
	188	{
	189	struct proc_dir_entry *pde = PDE(file_inode(file));
	190	loff_t rv = -EINVAL;
	191	if (use_pde(pde)) {
	192	loff_t (llseek)(struct file , loff_t, int);
	193	llseek = pde->proc_fops->llseek;
	194	if (!llseek)
	195	llseek = default_llseek;
	196	rv = llseek(file, offset, whence);
	197	unuse_pde(pde);
	198	}
786d7e16 AD	199	return rv;
	200	}
	201
866ad9a7	202	static ssize_t proc_reg_read(struct file file, char __user buf, size_t count, loff_t *ppos)
786d7e16	203	{
866ad9a7	204	ssize_t (read)(struct file , char __user , size_t, loff_t );
496ad9aa	205	struct proc_dir_entry *pde = PDE(file_inode(file));
786d7e16	206	ssize_t rv = -EIO;
866ad9a7 AV	207	if (use_pde(pde)) {
	208	read = pde->proc_fops->read;
	209	if (read)
	210	rv = read(file, buf, count, ppos);
	211	unuse_pde(pde);
786d7e16	212	}
866ad9a7 AV	213	return rv;
866ad9a7 AV	214	}
786d7e16	215
866ad9a7 AV	216	static ssize_t proc_reg_write(struct file file, const char __user buf, size_t count, loff_t *ppos)
	217	{
	218	ssize_t (write)(struct file , const char __user , size_t, loff_t );
	219	struct proc_dir_entry *pde = PDE(file_inode(file));
	220	ssize_t rv = -EIO;
	221	if (use_pde(pde)) {
	222	write = pde->proc_fops->write;
	223	if (write)
	224	rv = write(file, buf, count, ppos);
	225	unuse_pde(pde);
	226	}
786d7e16 AD	227	return rv;
	228	}
	229
	230	static unsigned int proc_reg_poll(struct file file, struct poll_table_struct pts)
	231	{
496ad9aa	232	struct proc_dir_entry *pde = PDE(file_inode(file));
dd23aae4	233	unsigned int rv = DEFAULT_POLLMASK;
786d7e16	234	unsigned int (poll)(struct file , struct poll_table_struct *);
866ad9a7 AV	235	if (use_pde(pde)) {
	236	poll = pde->proc_fops->poll;
	237	if (poll)
	238	rv = poll(file, pts);
	239	unuse_pde(pde);
786d7e16	240	}
786d7e16 AD	241	return rv;
	242	}
	243
	244	static long proc_reg_unlocked_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
	245	{
496ad9aa	246	struct proc_dir_entry *pde = PDE(file_inode(file));
786d7e16	247	long rv = -ENOTTY;
b19dd42f	248	long (ioctl)(struct file , unsigned int, unsigned long);
866ad9a7 AV	249	if (use_pde(pde)) {
	250	ioctl = pde->proc_fops->unlocked_ioctl;
	251	if (ioctl)
	252	rv = ioctl(file, cmd, arg);
	253	unuse_pde(pde);
786d7e16	254	}
786d7e16 AD	255	return rv;
	256	}
	257
	258	#ifdef CONFIG_COMPAT
	259	static long proc_reg_compat_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
	260	{
496ad9aa	261	struct proc_dir_entry *pde = PDE(file_inode(file));
786d7e16 AD	262	long rv = -ENOTTY;
786d7e16 AD	263	long (compat_ioctl)(struct file , unsigned int, unsigned long);
866ad9a7 AV	264	if (use_pde(pde)) {
	265	compat_ioctl = pde->proc_fops->compat_ioctl;
	266	if (compat_ioctl)
	267	rv = compat_ioctl(file, cmd, arg);
	268	unuse_pde(pde);
786d7e16	269	}
786d7e16 AD	270	return rv;
	271	}
	272	#endif
	273
	274	static int proc_reg_mmap(struct file file, struct vm_area_struct vma)
	275	{
496ad9aa	276	struct proc_dir_entry *pde = PDE(file_inode(file));
786d7e16 AD	277	int rv = -EIO;
786d7e16 AD	278	int (mmap)(struct file , struct vm_area_struct *);
866ad9a7 AV	279	if (use_pde(pde)) {
	280	mmap = pde->proc_fops->mmap;
	281	if (mmap)
	282	rv = mmap(file, vma);
	283	unuse_pde(pde);
786d7e16	284	}
786d7e16 AD	285	return rv;
	286	}
	287
	288	static int proc_reg_open(struct inode inode, struct file file)
	289	{
	290	struct proc_dir_entry *pde = PDE(inode);
	291	int rv = 0;
	292	int (open)(struct inode , struct file *);
881adb85 AD	293	int (release)(struct inode , struct file *);
	294	struct pde_opener *pdeo;
	295
	296	/*
	297	* What for, you ask? Well, we can have open, rmmod, remove_proc_entry
	298	* sequence. ->release won't be called because ->proc_fops will be
	299	* cleared. Depending on complexity of ->release, consequences vary.
	300	*
	301	* We can't wait for mercy when close will be done for real, it's
	302	* deadlockable: rmmod foo </proc/foo . So, we're going to do ->release
	303	* by hand in remove_proc_entry(). For this, save opener's credentials
	304	* for later.
	305	*/
05c0ae21	306	pdeo = kzalloc(sizeof(struct pde_opener), GFP_KERNEL);
881adb85 AD	307	if (!pdeo)
881adb85 AD	308	return -ENOMEM;
786d7e16	309
866ad9a7	310	if (!use_pde(pde)) {
881adb85	311	kfree(pdeo);
d2857e79	312	return -ENOENT;
786d7e16	313	}
786d7e16	314	open = pde->proc_fops->open;
881adb85	315	release = pde->proc_fops->release;
786d7e16 AD	316
	317	if (open)
	318	rv = open(inode, file);
	319
881adb85 AD	320	if (rv == 0 && release) {
881adb85 AD	321	/* To know what to release. */
881adb85 AD	322	pdeo->file = file;
881adb85 AD	323	/* Strictly for "too late" ->release in proc_reg_release(). */
05c0ae21	324	spin_lock(&pde->pde_unload_lock);
881adb85	325	list_add(&pdeo->lh, &pde->pde_openers);
05c0ae21	326	spin_unlock(&pde->pde_unload_lock);
881adb85 AD	327	} else
881adb85 AD	328	kfree(pdeo);
05c0ae21 AV	329
05c0ae21 AV	330	unuse_pde(pde);
786d7e16 AD	331	return rv;
	332	}
	333
	334	static int proc_reg_release(struct inode inode, struct file file)
	335	{
	336	struct proc_dir_entry *pde = PDE(inode);
881adb85	337	struct pde_opener *pdeo;
786d7e16	338	spin_lock(&pde->pde_unload_lock);
ca469f35 AV	339	list_for_each_entry(pdeo, &pde->pde_openers, lh) {
	340	if (pdeo->file == file) {
	341	close_pdeo(pde, pdeo);
	342	break;
	343	}
881adb85	344	}
786d7e16	345	spin_unlock(&pde->pde_unload_lock);
ca469f35	346	return 0;
786d7e16 AD	347	}
	348
	349	static const struct file_operations proc_reg_file_ops = {
	350	.llseek = proc_reg_llseek,
	351	.read = proc_reg_read,
	352	.write = proc_reg_write,
	353	.poll = proc_reg_poll,
	354	.unlocked_ioctl = proc_reg_unlocked_ioctl,
	355	#ifdef CONFIG_COMPAT
	356	.compat_ioctl = proc_reg_compat_ioctl,
	357	#endif
	358	.mmap = proc_reg_mmap,
	359	.open = proc_reg_open,
	360	.release = proc_reg_release,
	361	};
	362
778f3dd5 DM	363	#ifdef CONFIG_COMPAT
	364	static const struct file_operations proc_reg_file_ops_no_compat = {
	365	.llseek = proc_reg_llseek,
	366	.read = proc_reg_read,
	367	.write = proc_reg_write,
	368	.poll = proc_reg_poll,
	369	.unlocked_ioctl = proc_reg_unlocked_ioctl,
	370	.mmap = proc_reg_mmap,
	371	.open = proc_reg_open,
	372	.release = proc_reg_release,
	373	};
	374	#endif
	375
6d1b6e4e	376	struct inode proc_get_inode(struct super_block sb, struct proc_dir_entry *de)
1da177e4	377	{
51f0885e	378	struct inode *inode = new_inode_pseudo(sb);
1da177e4	379
51f0885e LT	380	if (inode) {
51f0885e LT	381	inode->i_ino = de->low_ino;
a1d4aebb	382	inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME;
a1d4aebb	383	PROC_I(inode)->pde = de;
5e971dce AD	384
	385	if (de->mode) {
	386	inode->i_mode = de->mode;
	387	inode->i_uid = de->uid;
	388	inode->i_gid = de->gid;
	389	}
	390	if (de->size)
	391	inode->i_size = de->size;
	392	if (de->nlink)
bfe86848	393	set_nlink(inode, de->nlink);
b6cdc731 AV	394	WARN_ON(!de->proc_iops);
b6cdc731 AV	395	inode->i_op = de->proc_iops;
5e971dce AD	396	if (de->proc_fops) {
5e971dce AD	397	if (S_ISREG(inode->i_mode)) {
778f3dd5	398	#ifdef CONFIG_COMPAT
5e971dce AD	399	if (!de->proc_fops->compat_ioctl)
	400	inode->i_fop =
	401	&proc_reg_file_ops_no_compat;
	402	else
778f3dd5	403	#endif
5e971dce AD	404	inode->i_fop = &proc_reg_file_ops;
	405	} else {
	406	inode->i_fop = de->proc_fops;
778f3dd5	407	}
786d7e16	408	}
99b76233	409	} else
135d5655	410	pde_put(de);
1da177e4	411	return inode;
d3d009cb	412	}
1da177e4	413
07543f5c	414	int proc_fill_super(struct super_block *s)
1da177e4	415	{
87e0aab3 MP	416	struct inode *root_inode;
87e0aab3 MP	417
92d03285	418	s->s_flags \|= MS_NODIRATIME \| MS_NOSUID \| MS_NOEXEC;
1da177e4 LT	419	s->s_blocksize = 1024;
	420	s->s_blocksize_bits = 10;
	421	s->s_magic = PROC_SUPER_MAGIC;
	422	s->s_op = &proc_sops;
	423	s->s_time_gran = 1;
	424
135d5655	425	pde_get(&proc_root);
87e0aab3 MP	426	root_inode = proc_get_inode(s, &proc_root);
87e0aab3 MP	427	if (!root_inode) {
87ebdc00	428	pr_err("proc_fill_super: get root inode failed\n");
87e0aab3 MP	429	return -ENOMEM;
87e0aab3 MP	430	}
1da177e4	431
87e0aab3 MP	432	s->s_root = d_make_root(root_inode);
87e0aab3 MP	433	if (!s->s_root) {
87ebdc00	434	pr_err("proc_fill_super: allocate dentry failed\n");
87e0aab3 MP	435	return -ENOMEM;
	436	}
	437
021ada7d	438	return proc_setup_self(s);
1da177e4	439	}