[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_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 const char *proc_follow_link(struct dentry *dentry, void **cookie)
{
	struct proc_dir_entry *pde = PDE(d_inode(dentry));
	if (unlikely(!use_pde(pde)))
		return ERR_PTR(-EINVAL);
	*cookie = pde;
	return pde->data;
}

static void proc_put_link(struct inode *unused, void *p)
{
	unuse_pde(p);
}

const struct inode_operations proc_link_inode_operations = {
	.readlink	= generic_readlink,
	.follow_link	= proc_follow_link,
	.put_link	= proc_put_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)
{
	struct inode *root_inode;
	int ret;

	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;
	}

	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\|
040b5c6f	98	SLAB_MEM_SPREAD\|SLAB_PANIC),
20c2df83	99	init_once);
1da177e4 LT	100	}
1da177e4 LT	101
97412950 VK	102	static int proc_show_options(struct seq_file seq, struct dentry root)
97412950 VK	103	{
0499680a VK	104	struct super_block *sb = root->d_sb;
	105	struct pid_namespace *pid = sb->s_fs_info;
	106
dcb0f222 EB	107	if (!gid_eq(pid->pid_gid, GLOBAL_ROOT_GID))
dcb0f222 EB	108	seq_printf(seq, ",gid=%u", from_kgid_munged(&init_user_ns, pid->pid_gid));
0499680a VK	109	if (pid->hide_pid != 0)
	110	seq_printf(seq, ",hidepid=%u", pid->hide_pid);
	111
97412950 VK	112	return 0;
	113	}
	114
ee9b6d61	115	static const struct super_operations proc_sops = {
1da177e4 LT	116	.alloc_inode = proc_alloc_inode,
1da177e4 LT	117	.destroy_inode = proc_destroy_inode,
1da177e4	118	.drop_inode = generic_delete_inode,
8267952b	119	.evict_inode = proc_evict_inode,
1da177e4	120	.statfs = simple_statfs,
97412950 VK	121	.remount_fs = proc_remount,
97412950 VK	122	.show_options = proc_show_options,
1da177e4 LT	123	};
1da177e4 LT	124
866ad9a7 AV	125	enum {BIAS = -1U<<31};
	126
	127	static inline int use_pde(struct proc_dir_entry *pde)
	128	{
05c0ae21	129	return atomic_inc_unless_negative(&pde->in_use);
881adb85 AD	130	}
881adb85 AD	131
866ad9a7	132	static void unuse_pde(struct proc_dir_entry *pde)
881adb85	133	{
05c0ae21 AV	134	if (atomic_dec_return(&pde->in_use) == BIAS)
05c0ae21 AV	135	complete(pde->pde_unload_completion);
786d7e16 AD	136	}
786d7e16 AD	137
ca469f35 AV	138	/* pde is locked */
	139	static void close_pdeo(struct proc_dir_entry pde, struct pde_opener pdeo)
	140	{
05c0ae21	141	if (pdeo->closing) {
ca469f35	142	/* somebody else is doing that, just wait */
05c0ae21 AV	143	DECLARE_COMPLETION_ONSTACK(c);
05c0ae21 AV	144	pdeo->c = &c;
ca469f35	145	spin_unlock(&pde->pde_unload_lock);
05c0ae21	146	wait_for_completion(&c);
ca469f35	147	spin_lock(&pde->pde_unload_lock);
ca469f35 AV	148	} else {
ca469f35 AV	149	struct file *file;
05c0ae21	150	pdeo->closing = 1;
ca469f35 AV	151	spin_unlock(&pde->pde_unload_lock);
	152	file = pdeo->file;
	153	pde->proc_fops->release(file_inode(file), file);
	154	spin_lock(&pde->pde_unload_lock);
	155	list_del_init(&pdeo->lh);
05c0ae21 AV	156	if (pdeo->c)
05c0ae21 AV	157	complete(pdeo->c);
ca469f35	158	kfree(pdeo);
05c0ae21	159	}
ca469f35 AV	160	}
ca469f35 AV	161
866ad9a7	162	void proc_entry_rundown(struct proc_dir_entry *de)
786d7e16	163	{
05c0ae21	164	DECLARE_COMPLETION_ONSTACK(c);
866ad9a7	165	/* Wait until all existing callers into module are done. */
05c0ae21 AV	166	de->pde_unload_completion = &c;
	167	if (atomic_add_return(BIAS, &de->in_use) != BIAS)
	168	wait_for_completion(&c);
786d7e16	169
05c0ae21	170	spin_lock(&de->pde_unload_lock);
866ad9a7 AV	171	while (!list_empty(&de->pde_openers)) {
866ad9a7 AV	172	struct pde_opener *pdeo;
866ad9a7	173	pdeo = list_first_entry(&de->pde_openers, struct pde_opener, lh);
ca469f35	174	close_pdeo(de, pdeo);
866ad9a7 AV	175	}
866ad9a7 AV	176	spin_unlock(&de->pde_unload_lock);
786d7e16 AD	177	}
786d7e16 AD	178
866ad9a7 AV	179	static loff_t proc_reg_llseek(struct file *file, loff_t offset, int whence)
	180	{
	181	struct proc_dir_entry *pde = PDE(file_inode(file));
	182	loff_t rv = -EINVAL;
	183	if (use_pde(pde)) {
	184	loff_t (llseek)(struct file , loff_t, int);
	185	llseek = pde->proc_fops->llseek;
	186	if (!llseek)
	187	llseek = default_llseek;
	188	rv = llseek(file, offset, whence);
	189	unuse_pde(pde);
	190	}
786d7e16 AD	191	return rv;
	192	}
	193
866ad9a7	194	static ssize_t proc_reg_read(struct file file, char __user buf, size_t count, loff_t *ppos)
786d7e16	195	{
866ad9a7	196	ssize_t (read)(struct file , char __user , size_t, loff_t );
496ad9aa	197	struct proc_dir_entry *pde = PDE(file_inode(file));
786d7e16	198	ssize_t rv = -EIO;
866ad9a7 AV	199	if (use_pde(pde)) {
	200	read = pde->proc_fops->read;
	201	if (read)
	202	rv = read(file, buf, count, ppos);
	203	unuse_pde(pde);
786d7e16	204	}
866ad9a7 AV	205	return rv;
866ad9a7 AV	206	}
786d7e16	207
866ad9a7 AV	208	static ssize_t proc_reg_write(struct file file, const char __user buf, size_t count, loff_t *ppos)
	209	{
	210	ssize_t (write)(struct file , const char __user , size_t, loff_t );
	211	struct proc_dir_entry *pde = PDE(file_inode(file));
	212	ssize_t rv = -EIO;
	213	if (use_pde(pde)) {
	214	write = pde->proc_fops->write;
	215	if (write)
	216	rv = write(file, buf, count, ppos);
	217	unuse_pde(pde);
	218	}
786d7e16 AD	219	return rv;
	220	}
	221
	222	static unsigned int proc_reg_poll(struct file file, struct poll_table_struct pts)
	223	{
496ad9aa	224	struct proc_dir_entry *pde = PDE(file_inode(file));
dd23aae4	225	unsigned int rv = DEFAULT_POLLMASK;
786d7e16	226	unsigned int (poll)(struct file , struct poll_table_struct *);
866ad9a7 AV	227	if (use_pde(pde)) {
	228	poll = pde->proc_fops->poll;
	229	if (poll)
	230	rv = poll(file, pts);
	231	unuse_pde(pde);
786d7e16	232	}
786d7e16 AD	233	return rv;
	234	}
	235
	236	static long proc_reg_unlocked_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
	237	{
496ad9aa	238	struct proc_dir_entry *pde = PDE(file_inode(file));
786d7e16	239	long rv = -ENOTTY;
b19dd42f	240	long (ioctl)(struct file , unsigned int, unsigned long);
866ad9a7 AV	241	if (use_pde(pde)) {
	242	ioctl = pde->proc_fops->unlocked_ioctl;
	243	if (ioctl)
	244	rv = ioctl(file, cmd, arg);
	245	unuse_pde(pde);
786d7e16	246	}
786d7e16 AD	247	return rv;
	248	}
	249
	250	#ifdef CONFIG_COMPAT
	251	static long proc_reg_compat_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
	252	{
496ad9aa	253	struct proc_dir_entry *pde = PDE(file_inode(file));
786d7e16 AD	254	long rv = -ENOTTY;
786d7e16 AD	255	long (compat_ioctl)(struct file , unsigned int, unsigned long);
866ad9a7 AV	256	if (use_pde(pde)) {
	257	compat_ioctl = pde->proc_fops->compat_ioctl;
	258	if (compat_ioctl)
	259	rv = compat_ioctl(file, cmd, arg);
	260	unuse_pde(pde);
786d7e16	261	}
786d7e16 AD	262	return rv;
	263	}
	264	#endif
	265
	266	static int proc_reg_mmap(struct file file, struct vm_area_struct vma)
	267	{
496ad9aa	268	struct proc_dir_entry *pde = PDE(file_inode(file));
786d7e16 AD	269	int rv = -EIO;
786d7e16 AD	270	int (mmap)(struct file , struct vm_area_struct *);
866ad9a7 AV	271	if (use_pde(pde)) {
	272	mmap = pde->proc_fops->mmap;
	273	if (mmap)
	274	rv = mmap(file, vma);
	275	unuse_pde(pde);
786d7e16	276	}
786d7e16 AD	277	return rv;
	278	}
	279
5721cf84 HD	280	static unsigned long
	281	proc_reg_get_unmapped_area(struct file *file, unsigned long orig_addr,
	282	unsigned long len, unsigned long pgoff,
	283	unsigned long flags)
c4fe2448 AD	284	{
c4fe2448 AD	285	struct proc_dir_entry *pde = PDE(file_inode(file));
2cbe3b0a	286	unsigned long rv = -EIO;
ae5758a1	287
c4fe2448	288	if (use_pde(pde)) {
ae5758a1 JB	289	typeof(proc_reg_get_unmapped_area) *get_area;
	290
	291	get_area = pde->proc_fops->get_unmapped_area;
fad1a86e	292	#ifdef CONFIG_MMU
ae5758a1 JB	293	if (!get_area)
ae5758a1 JB	294	get_area = current->mm->get_unmapped_area;
fad1a86e	295	#endif
ae5758a1	296
5721cf84 HD	297	if (get_area)
5721cf84 HD	298	rv = get_area(file, orig_addr, len, pgoff, flags);
ae5758a1 JB	299	else
ae5758a1 JB	300	rv = orig_addr;
c4fe2448 AD	301	unuse_pde(pde);
	302	}
	303	return rv;
	304	}
	305
786d7e16 AD	306	static int proc_reg_open(struct inode inode, struct file file)
	307	{
	308	struct proc_dir_entry *pde = PDE(inode);
	309	int rv = 0;
	310	int (open)(struct inode , struct file *);
881adb85 AD	311	int (release)(struct inode , struct file *);
	312	struct pde_opener *pdeo;
	313
	314	/*
	315	* What for, you ask? Well, we can have open, rmmod, remove_proc_entry
	316	* sequence. ->release won't be called because ->proc_fops will be
	317	* cleared. Depending on complexity of ->release, consequences vary.
	318	*
	319	* We can't wait for mercy when close will be done for real, it's
	320	* deadlockable: rmmod foo </proc/foo . So, we're going to do ->release
	321	* by hand in remove_proc_entry(). For this, save opener's credentials
	322	* for later.
	323	*/
05c0ae21	324	pdeo = kzalloc(sizeof(struct pde_opener), GFP_KERNEL);
881adb85 AD	325	if (!pdeo)
881adb85 AD	326	return -ENOMEM;
786d7e16	327
866ad9a7	328	if (!use_pde(pde)) {
881adb85	329	kfree(pdeo);
d2857e79	330	return -ENOENT;
786d7e16	331	}
786d7e16	332	open = pde->proc_fops->open;
881adb85	333	release = pde->proc_fops->release;
786d7e16 AD	334
	335	if (open)
	336	rv = open(inode, file);
	337
881adb85 AD	338	if (rv == 0 && release) {
881adb85 AD	339	/* To know what to release. */
881adb85 AD	340	pdeo->file = file;
881adb85 AD	341	/* Strictly for "too late" ->release in proc_reg_release(). */
05c0ae21	342	spin_lock(&pde->pde_unload_lock);
881adb85	343	list_add(&pdeo->lh, &pde->pde_openers);
05c0ae21	344	spin_unlock(&pde->pde_unload_lock);
881adb85 AD	345	} else
881adb85 AD	346	kfree(pdeo);
05c0ae21 AV	347
05c0ae21 AV	348	unuse_pde(pde);
786d7e16 AD	349	return rv;
	350	}
	351
	352	static int proc_reg_release(struct inode inode, struct file file)
	353	{
	354	struct proc_dir_entry *pde = PDE(inode);
881adb85	355	struct pde_opener *pdeo;
786d7e16	356	spin_lock(&pde->pde_unload_lock);
ca469f35 AV	357	list_for_each_entry(pdeo, &pde->pde_openers, lh) {
	358	if (pdeo->file == file) {
	359	close_pdeo(pde, pdeo);
	360	break;
	361	}
881adb85	362	}
786d7e16	363	spin_unlock(&pde->pde_unload_lock);
ca469f35	364	return 0;
786d7e16 AD	365	}
	366
	367	static const struct file_operations proc_reg_file_ops = {
	368	.llseek = proc_reg_llseek,
	369	.read = proc_reg_read,
	370	.write = proc_reg_write,
	371	.poll = proc_reg_poll,
	372	.unlocked_ioctl = proc_reg_unlocked_ioctl,
	373	#ifdef CONFIG_COMPAT
	374	.compat_ioctl = proc_reg_compat_ioctl,
	375	#endif
	376	.mmap = proc_reg_mmap,
c4fe2448	377	.get_unmapped_area = proc_reg_get_unmapped_area,
786d7e16 AD	378	.open = proc_reg_open,
	379	.release = proc_reg_release,
	380	};
	381
778f3dd5 DM	382	#ifdef CONFIG_COMPAT
	383	static const struct file_operations proc_reg_file_ops_no_compat = {
	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	.mmap = proc_reg_mmap,
c4fe2448	390	.get_unmapped_area = proc_reg_get_unmapped_area,
778f3dd5 DM	391	.open = proc_reg_open,
	392	.release = proc_reg_release,
	393	};
	394	#endif
	395
6e77137b	396	static const char proc_follow_link(struct dentry dentry, void **cookie)
7e0e953b	397	{
2b0143b5	398	struct proc_dir_entry *pde = PDE(d_inode(dentry));
7e0e953b AV	399	if (unlikely(!use_pde(pde)))
7e0e953b AV	400	return ERR_PTR(-EINVAL);
680baacb AV	401	*cookie = pde;
680baacb AV	402	return pde->data;
7e0e953b AV	403	}
7e0e953b AV	404
5f2c4179	405	static void proc_put_link(struct inode unused, void p)
7e0e953b AV	406	{
	407	unuse_pde(p);
	408	}
	409
	410	const struct inode_operations proc_link_inode_operations = {
	411	.readlink = generic_readlink,
	412	.follow_link = proc_follow_link,
	413	.put_link = proc_put_link,
	414	};
	415
6d1b6e4e	416	struct inode proc_get_inode(struct super_block sb, struct proc_dir_entry *de)
1da177e4	417	{
51f0885e	418	struct inode *inode = new_inode_pseudo(sb);
1da177e4	419
51f0885e LT	420	if (inode) {
51f0885e LT	421	inode->i_ino = de->low_ino;
a1d4aebb	422	inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME;
a1d4aebb	423	PROC_I(inode)->pde = de;
5e971dce	424
eb6d38d5 EB	425	if (is_empty_pde(de)) {
	426	make_empty_dir_inode(inode);
	427	return inode;
	428	}
5e971dce AD	429	if (de->mode) {
	430	inode->i_mode = de->mode;
	431	inode->i_uid = de->uid;
	432	inode->i_gid = de->gid;
	433	}
	434	if (de->size)
	435	inode->i_size = de->size;
	436	if (de->nlink)
bfe86848	437	set_nlink(inode, de->nlink);
b6cdc731 AV	438	WARN_ON(!de->proc_iops);
b6cdc731 AV	439	inode->i_op = de->proc_iops;
5e971dce AD	440	if (de->proc_fops) {
5e971dce AD	441	if (S_ISREG(inode->i_mode)) {
778f3dd5	442	#ifdef CONFIG_COMPAT
5e971dce AD	443	if (!de->proc_fops->compat_ioctl)
	444	inode->i_fop =
	445	&proc_reg_file_ops_no_compat;
	446	else
778f3dd5	447	#endif
5e971dce AD	448	inode->i_fop = &proc_reg_file_ops;
	449	} else {
	450	inode->i_fop = de->proc_fops;
778f3dd5	451	}
786d7e16	452	}
99b76233	453	} else
135d5655	454	pde_put(de);
1da177e4	455	return inode;
d3d009cb	456	}
1da177e4	457
07543f5c	458	int proc_fill_super(struct super_block *s)
1da177e4	459	{
87e0aab3	460	struct inode *root_inode;
0097875b	461	int ret;
87e0aab3	462
92d03285	463	s->s_flags \|= MS_NODIRATIME \| MS_NOSUID \| MS_NOEXEC;
1da177e4 LT	464	s->s_blocksize = 1024;
	465	s->s_blocksize_bits = 10;
	466	s->s_magic = PROC_SUPER_MAGIC;
	467	s->s_op = &proc_sops;
	468	s->s_time_gran = 1;
	469
135d5655	470	pde_get(&proc_root);
87e0aab3 MP	471	root_inode = proc_get_inode(s, &proc_root);
87e0aab3 MP	472	if (!root_inode) {
87ebdc00	473	pr_err("proc_fill_super: get root inode failed\n");
87e0aab3 MP	474	return -ENOMEM;
87e0aab3 MP	475	}
1da177e4	476
87e0aab3 MP	477	s->s_root = d_make_root(root_inode);
87e0aab3 MP	478	if (!s->s_root) {
87ebdc00	479	pr_err("proc_fill_super: allocate dentry failed\n");
87e0aab3 MP	480	return -ENOMEM;
	481	}
	482
0097875b EB	483	ret = proc_setup_self(s);
	484	if (ret) {
	485	return ret;
	486	}
	487	return proc_setup_thread_self(s);
1da177e4	488	}