[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/mm.h>
#include <linux/string.h>
#include <linux/stat.h>
#include <linux/completion.h>
#include <linux/poll.h>
#include <linux/file.h>
#include <linux/limits.h>
#include <linux/init.h>
#include <linux/module.h>
#include <linux/smp_lock.h>
#include <linux/sysctl.h>
#include <linux/slab.h>

#include <asm/system.h>
#include <asm/uaccess.h>

#include "internal.h"

static void proc_evict_inode(struct inode *inode)
{
	struct proc_dir_entry *de;

	truncate_inode_pages(&inode->i_data, 0);
	end_writeback(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);
	if (PROC_I(inode)->sysctl)
		sysctl_head_put(PROC_I(inode)->sysctl);
}

struct vfsmount *proc_mnt;

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;
	inode = &ei->vfs_inode;
	inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME;
	return inode;
}

static void proc_destroy_inode(struct inode *inode)
{
	kmem_cache_free(proc_inode_cachep, PROC_I(inode));
}

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 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,
};

static void __pde_users_dec(struct proc_dir_entry *pde)
{
	pde->pde_users--;
	if (pde->pde_unload_completion && pde->pde_users == 0)
		complete(pde->pde_unload_completion);
}

void pde_users_dec(struct proc_dir_entry *pde)
{
	spin_lock(&pde->pde_unload_lock);
	__pde_users_dec(pde);
	spin_unlock(&pde->pde_unload_lock);
}

static loff_t proc_reg_llseek(struct file *file, loff_t offset, int whence)
{
	struct proc_dir_entry *pde = PDE(file->f_path.dentry->d_inode);
	loff_t rv = -EINVAL;
	loff_t (*llseek)(struct file *, loff_t, int);

	spin_lock(&pde->pde_unload_lock);
	/*
	 * remove_proc_entry() is going to delete PDE (as part of module
	 * cleanup sequence). No new callers into module allowed.
	 */
	if (!pde->proc_fops) {
		spin_unlock(&pde->pde_unload_lock);
		return rv;
	}
	/*
	 * Bump refcount so that remove_proc_entry will wail for ->llseek to
	 * complete.
	 */
	pde->pde_users++;
	/*
	 * Save function pointer under lock, to protect against ->proc_fops
	 * NULL'ifying right after ->pde_unload_lock is dropped.
	 */
	llseek = pde->proc_fops->llseek;
	spin_unlock(&pde->pde_unload_lock);

	if (!llseek)
		llseek = default_llseek;
	rv = llseek(file, offset, whence);

	pde_users_dec(pde);
	return rv;
}

static ssize_t proc_reg_read(struct file *file, char __user *buf, size_t count, loff_t *ppos)
{
	struct proc_dir_entry *pde = PDE(file->f_path.dentry->d_inode);
	ssize_t rv = -EIO;
	ssize_t (*read)(struct file *, char __user *, size_t, loff_t *);

	spin_lock(&pde->pde_unload_lock);
	if (!pde->proc_fops) {
		spin_unlock(&pde->pde_unload_lock);
		return rv;
	}
	pde->pde_users++;
	read = pde->proc_fops->read;
	spin_unlock(&pde->pde_unload_lock);

	if (read)
		rv = read(file, buf, count, ppos);

	pde_users_dec(pde);
	return rv;
}

static ssize_t proc_reg_write(struct file *file, const char __user *buf, size_t count, loff_t *ppos)
{
	struct proc_dir_entry *pde = PDE(file->f_path.dentry->d_inode);
	ssize_t rv = -EIO;
	ssize_t (*write)(struct file *, const char __user *, size_t, loff_t *);

	spin_lock(&pde->pde_unload_lock);
	if (!pde->proc_fops) {
		spin_unlock(&pde->pde_unload_lock);
		return rv;
	}
	pde->pde_users++;
	write = pde->proc_fops->write;
	spin_unlock(&pde->pde_unload_lock);

	if (write)
		rv = write(file, buf, count, ppos);

	pde_users_dec(pde);
	return rv;
}

static unsigned int proc_reg_poll(struct file *file, struct poll_table_struct *pts)
{
	struct proc_dir_entry *pde = PDE(file->f_path.dentry->d_inode);
	unsigned int rv = DEFAULT_POLLMASK;
	unsigned int (*poll)(struct file *, struct poll_table_struct *);

	spin_lock(&pde->pde_unload_lock);
	if (!pde->proc_fops) {
		spin_unlock(&pde->pde_unload_lock);
		return rv;
	}
	pde->pde_users++;
	poll = pde->proc_fops->poll;
	spin_unlock(&pde->pde_unload_lock);

	if (poll)
		rv = poll(file, pts);

	pde_users_dec(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->f_path.dentry->d_inode);
	long rv = -ENOTTY;
	long (*unlocked_ioctl)(struct file *, unsigned int, unsigned long);
	int (*ioctl)(struct inode *, struct file *, unsigned int, unsigned long);

	spin_lock(&pde->pde_unload_lock);
	if (!pde->proc_fops) {
		spin_unlock(&pde->pde_unload_lock);
		return rv;
	}
	pde->pde_users++;
	unlocked_ioctl = pde->proc_fops->unlocked_ioctl;
	ioctl = pde->proc_fops->ioctl;
	spin_unlock(&pde->pde_unload_lock);

	if (unlocked_ioctl) {
		rv = unlocked_ioctl(file, cmd, arg);
		if (rv == -ENOIOCTLCMD)
			rv = -EINVAL;
	} else if (ioctl) {
		WARN_ONCE(1, "Procfs ioctl handlers must use unlocked_ioctl, "
			  "%pf will be called without the Bkl held\n", ioctl);
		rv = ioctl(file->f_path.dentry->d_inode, file, cmd, arg);
	}

	pde_users_dec(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->f_path.dentry->d_inode);
	long rv = -ENOTTY;
	long (*compat_ioctl)(struct file *, unsigned int, unsigned long);

	spin_lock(&pde->pde_unload_lock);
	if (!pde->proc_fops) {
		spin_unlock(&pde->pde_unload_lock);
		return rv;
	}
	pde->pde_users++;
	compat_ioctl = pde->proc_fops->compat_ioctl;
	spin_unlock(&pde->pde_unload_lock);

	if (compat_ioctl)
		rv = compat_ioctl(file, cmd, arg);

	pde_users_dec(pde);
	return rv;
}
#endif

static int proc_reg_mmap(struct file *file, struct vm_area_struct *vma)
{
	struct proc_dir_entry *pde = PDE(file->f_path.dentry->d_inode);
	int rv = -EIO;
	int (*mmap)(struct file *, struct vm_area_struct *);

	spin_lock(&pde->pde_unload_lock);
	if (!pde->proc_fops) {
		spin_unlock(&pde->pde_unload_lock);
		return rv;
	}
	pde->pde_users++;
	mmap = pde->proc_fops->mmap;
	spin_unlock(&pde->pde_unload_lock);

	if (mmap)
		rv = mmap(file, vma);

	pde_users_dec(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 = kmalloc(sizeof(struct pde_opener), GFP_KERNEL);
	if (!pdeo)
		return -ENOMEM;

	spin_lock(&pde->pde_unload_lock);
	if (!pde->proc_fops) {
		spin_unlock(&pde->pde_unload_lock);
		kfree(pdeo);
		return -EINVAL;
	}
	pde->pde_users++;
	open = pde->proc_fops->open;
	release = pde->proc_fops->release;
	spin_unlock(&pde->pde_unload_lock);

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

	spin_lock(&pde->pde_unload_lock);
	if (rv == 0 && release) {
		/* To know what to release. */
		pdeo->inode = inode;
		pdeo->file = file;
		/* Strictly for "too late" ->release in proc_reg_release(). */
		pdeo->release = release;
		list_add(&pdeo->lh, &pde->pde_openers);
	} else
		kfree(pdeo);
	__pde_users_dec(pde);
	spin_unlock(&pde->pde_unload_lock);
	return rv;
}

static struct pde_opener *find_pde_opener(struct proc_dir_entry *pde,
					struct inode *inode, struct file *file)
{
	struct pde_opener *pdeo;

	list_for_each_entry(pdeo, &pde->pde_openers, lh) {
		if (pdeo->inode == inode && pdeo->file == file)
			return pdeo;
	}
	return NULL;
}

static int proc_reg_release(struct inode *inode, struct file *file)
{
	struct proc_dir_entry *pde = PDE(inode);
	int rv = 0;
	int (*release)(struct inode *, struct file *);
	struct pde_opener *pdeo;

	spin_lock(&pde->pde_unload_lock);
	pdeo = find_pde_opener(pde, inode, file);
	if (!pde->proc_fops) {
		/*
		 * Can't simply exit, __fput() will think that everything is OK,
		 * and move on to freeing struct file. remove_proc_entry() will
		 * find slacker in opener's list and will try to do non-trivial
		 * things with struct file. Therefore, remove opener from list.
		 *
		 * But if opener is removed from list, who will ->release it?
		 */
		if (pdeo) {
			list_del(&pdeo->lh);
			spin_unlock(&pde->pde_unload_lock);
			rv = pdeo->release(inode, file);
			kfree(pdeo);
		} else
			spin_unlock(&pde->pde_unload_lock);
		return rv;
	}
	pde->pde_users++;
	release = pde->proc_fops->release;
	if (pdeo) {
		list_del(&pdeo->lh);
		kfree(pdeo);
	}
	spin_unlock(&pde->pde_unload_lock);

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

	pde_users_dec(pde);
	return rv;
}

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, unsigned int ino,
				struct proc_dir_entry *de)
{
	struct inode * inode;

	inode = iget_locked(sb, ino);
	if (!inode)
		return NULL;
	if (inode->i_state & I_NEW) {
		inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME;
		PROC_I(inode)->fd = 0;
		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)
			inode->i_nlink = de->nlink;
		if (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;
			}
		}
		unlock_new_inode(inode);
	} 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_INO, &proc_root);
	if (!root_inode)
		goto out_no_root;
	root_inode->i_uid = 0;
	root_inode->i_gid = 0;
	s->s_root = d_alloc_root(root_inode);
	if (!s->s_root)
		goto out_no_root;
	return 0;

out_no_root:
	printk("proc_read_super: get root inode failed\n");
	iput(root_inode);
	pde_put(&proc_root);
	return -ENOMEM;
}
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>
	10	#include <linux/mm.h>
	11	#include <linux/string.h>
	12	#include <linux/stat.h>
786d7e16	13	#include <linux/completion.h>
dd23aae4	14	#include <linux/poll.h>
1da177e4 LT	15	#include <linux/file.h>
	16	#include <linux/limits.h>
	17	#include <linux/init.h>
	18	#include <linux/module.h>
	19	#include <linux/smp_lock.h>
9043476f	20	#include <linux/sysctl.h>
5a0e3ad6	21	#include <linux/slab.h>
1da177e4 LT	22
	23	#include <asm/system.h>
	24	#include <asm/uaccess.h>
	25
fee781e6	26	#include "internal.h"
1da177e4	27
8267952b	28	static void proc_evict_inode(struct inode *inode)
1da177e4 LT	29	{
1da177e4 LT	30	struct proc_dir_entry *de;
1da177e4	31
fef26658	32	truncate_inode_pages(&inode->i_data, 0);
8267952b	33	end_writeback(inode);
fef26658	34
99f89551	35	/* Stop tracking associated processes */
13b41b09	36	put_pid(PROC_I(inode)->pid);
1da177e4 LT	37
	38	/* Let go of any associated proc directory entry */
	39	de = PROC_I(inode)->pde;
99b76233	40	if (de)
135d5655	41	pde_put(de);
9043476f AV	42	if (PROC_I(inode)->sysctl)
9043476f AV	43	sysctl_head_put(PROC_I(inode)->sysctl);
1da177e4 LT	44	}
	45
	46	struct vfsmount *proc_mnt;
	47
e18b890b	48	static struct kmem_cache * proc_inode_cachep;
1da177e4 LT	49
	50	static struct inode proc_alloc_inode(struct super_block sb)
	51	{
	52	struct proc_inode *ei;
	53	struct inode *inode;
	54
e94b1766	55	ei = (struct proc_inode *)kmem_cache_alloc(proc_inode_cachep, GFP_KERNEL);
1da177e4 LT	56	if (!ei)
1da177e4 LT	57	return NULL;
13b41b09	58	ei->pid = NULL;
aed7a6c4	59	ei->fd = 0;
1da177e4 LT	60	ei->op.proc_get_link = NULL;
1da177e4 LT	61	ei->pde = NULL;
9043476f AV	62	ei->sysctl = NULL;
9043476f AV	63	ei->sysctl_entry = NULL;
1da177e4 LT	64	inode = &ei->vfs_inode;
	65	inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME;
	66	return inode;
	67	}
	68
	69	static void proc_destroy_inode(struct inode *inode)
	70	{
	71	kmem_cache_free(proc_inode_cachep, PROC_I(inode));
	72	}
	73
51cc5068	74	static void init_once(void *foo)
1da177e4 LT	75	{
	76	struct proc_inode ei = (struct proc_inode ) foo;
	77
a35afb83	78	inode_init_once(&ei->vfs_inode);
1da177e4	79	}
20c2df83	80
5bcd7ff9	81	void __init proc_init_inodecache(void)
1da177e4 LT	82	{
	83	proc_inode_cachep = kmem_cache_create("proc_inode_cache",
	84	sizeof(struct proc_inode),
fffb60f9	85	0, (SLAB_RECLAIM_ACCOUNT\|
040b5c6f	86	SLAB_MEM_SPREAD\|SLAB_PANIC),
20c2df83	87	init_once);
1da177e4 LT	88	}
1da177e4 LT	89
ee9b6d61	90	static const struct super_operations proc_sops = {
1da177e4 LT	91	.alloc_inode = proc_alloc_inode,
1da177e4 LT	92	.destroy_inode = proc_destroy_inode,
1da177e4	93	.drop_inode = generic_delete_inode,
8267952b	94	.evict_inode = proc_evict_inode,
1da177e4	95	.statfs = simple_statfs,
1da177e4 LT	96	};
1da177e4 LT	97
881adb85	98	static void __pde_users_dec(struct proc_dir_entry *pde)
786d7e16	99	{
786d7e16 AD	100	pde->pde_users--;
	101	if (pde->pde_unload_completion && pde->pde_users == 0)
	102	complete(pde->pde_unload_completion);
881adb85 AD	103	}
881adb85 AD	104
3dec7f59	105	void pde_users_dec(struct proc_dir_entry *pde)
881adb85 AD	106	{
	107	spin_lock(&pde->pde_unload_lock);
	108	__pde_users_dec(pde);
786d7e16 AD	109	spin_unlock(&pde->pde_unload_lock);
	110	}
	111
	112	static loff_t proc_reg_llseek(struct file *file, loff_t offset, int whence)
	113	{
	114	struct proc_dir_entry *pde = PDE(file->f_path.dentry->d_inode);
	115	loff_t rv = -EINVAL;
	116	loff_t (llseek)(struct file , loff_t, int);
	117
	118	spin_lock(&pde->pde_unload_lock);
	119	/*
	120	* remove_proc_entry() is going to delete PDE (as part of module
	121	* cleanup sequence). No new callers into module allowed.
	122	*/
	123	if (!pde->proc_fops) {
	124	spin_unlock(&pde->pde_unload_lock);
	125	return rv;
	126	}
	127	/*
	128	* Bump refcount so that remove_proc_entry will wail for ->llseek to
	129	* complete.
	130	*/
	131	pde->pde_users++;
	132	/*
	133	* Save function pointer under lock, to protect against ->proc_fops
	134	* NULL'ifying right after ->pde_unload_lock is dropped.
	135	*/
	136	llseek = pde->proc_fops->llseek;
	137	spin_unlock(&pde->pde_unload_lock);
	138
	139	if (!llseek)
	140	llseek = default_llseek;
	141	rv = llseek(file, offset, whence);
	142
	143	pde_users_dec(pde);
	144	return rv;
	145	}
	146
	147	static ssize_t proc_reg_read(struct file file, char __user buf, size_t count, loff_t *ppos)
	148	{
	149	struct proc_dir_entry *pde = PDE(file->f_path.dentry->d_inode);
	150	ssize_t rv = -EIO;
	151	ssize_t (read)(struct file , char __user , size_t, loff_t );
	152
	153	spin_lock(&pde->pde_unload_lock);
	154	if (!pde->proc_fops) {
	155	spin_unlock(&pde->pde_unload_lock);
	156	return rv;
	157	}
	158	pde->pde_users++;
	159	read = pde->proc_fops->read;
	160	spin_unlock(&pde->pde_unload_lock);
	161
	162	if (read)
	163	rv = read(file, buf, count, ppos);
	164
	165	pde_users_dec(pde);
	166	return rv;
	167	}
	168
	169	static ssize_t proc_reg_write(struct file file, const char __user buf, size_t count, loff_t *ppos)
	170	{
	171	struct proc_dir_entry *pde = PDE(file->f_path.dentry->d_inode);
	172	ssize_t rv = -EIO;
173	ssize_t (write)(struct file , const char __user , size_t, loff_t );
174
175	spin_lock(&pde->pde_unload_lock);
176	if (!pde->proc_fops) {
177	spin_unlock(&pde->pde_unload_lock);
178	return rv;
179	}
180	pde->pde_users++;
181	write = pde->proc_fops->write;
182	spin_unlock(&pde->pde_unload_lock);
183
184	if (write)
185	rv = write(file, buf, count, ppos);
186
187	pde_users_dec(pde);
188	return rv;
189	}
190
191	static unsigned int proc_reg_poll(struct file file, struct poll_table_struct pts)
192	{
193	struct proc_dir_entry *pde = PDE(file->f_path.dentry->d_inode);
dd23aae4	194	unsigned int rv = DEFAULT_POLLMASK;
786d7e16 AD	195	unsigned int (poll)(struct file , struct poll_table_struct *);
	196
	197	spin_lock(&pde->pde_unload_lock);
	198	if (!pde->proc_fops) {
	199	spin_unlock(&pde->pde_unload_lock);
	200	return rv;
	201	}
	202	pde->pde_users++;
	203	poll = pde->proc_fops->poll;
	204	spin_unlock(&pde->pde_unload_lock);
	205
	206	if (poll)
	207	rv = poll(file, pts);
	208
	209	pde_users_dec(pde);
	210	return rv;
	211	}
	212
	213	static long proc_reg_unlocked_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
	214	{
	215	struct proc_dir_entry *pde = PDE(file->f_path.dentry->d_inode);
	216	long rv = -ENOTTY;
	217	long (unlocked_ioctl)(struct file , unsigned int, unsigned long);
	218	int (ioctl)(struct inode , struct file *, unsigned int, unsigned long);
	219
	220	spin_lock(&pde->pde_unload_lock);
	221	if (!pde->proc_fops) {
	222	spin_unlock(&pde->pde_unload_lock);
	223	return rv;
	224	}
	225	pde->pde_users++;
	226	unlocked_ioctl = pde->proc_fops->unlocked_ioctl;
	227	ioctl = pde->proc_fops->ioctl;
	228	spin_unlock(&pde->pde_unload_lock);
	229
	230	if (unlocked_ioctl) {
	231	rv = unlocked_ioctl(file, cmd, arg);
	232	if (rv == -ENOIOCTLCMD)
	233	rv = -EINVAL;
	234	} else if (ioctl) {
c2f98050 FW	235	WARN_ONCE(1, "Procfs ioctl handlers must use unlocked_ioctl, "
c2f98050 FW	236	"%pf will be called without the Bkl held\n", ioctl);
786d7e16	237	rv = ioctl(file->f_path.dentry->d_inode, file, cmd, arg);
786d7e16 AD	238	}
	239
	240	pde_users_dec(pde);
	241	return rv;
	242	}
	243
	244	#ifdef CONFIG_COMPAT
	245	static long proc_reg_compat_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
	246	{
	247	struct proc_dir_entry *pde = PDE(file->f_path.dentry->d_inode);
	248	long rv = -ENOTTY;
	249	long (compat_ioctl)(struct file , unsigned int, unsigned long);
	250
	251	spin_lock(&pde->pde_unload_lock);
	252	if (!pde->proc_fops) {
	253	spin_unlock(&pde->pde_unload_lock);
	254	return rv;
	255	}
	256	pde->pde_users++;
	257	compat_ioctl = pde->proc_fops->compat_ioctl;
	258	spin_unlock(&pde->pde_unload_lock);
	259
	260	if (compat_ioctl)
	261	rv = compat_ioctl(file, cmd, arg);
	262
	263	pde_users_dec(pde);
	264	return rv;
	265	}
	266	#endif
	267
	268	static int proc_reg_mmap(struct file file, struct vm_area_struct vma)
	269	{
	270	struct proc_dir_entry *pde = PDE(file->f_path.dentry->d_inode);
	271	int rv = -EIO;
	272	int (mmap)(struct file , struct vm_area_struct *);
	273
	274	spin_lock(&pde->pde_unload_lock);
	275	if (!pde->proc_fops) {
	276	spin_unlock(&pde->pde_unload_lock);
	277	return rv;
	278	}
	279	pde->pde_users++;
	280	mmap = pde->proc_fops->mmap;
	281	spin_unlock(&pde->pde_unload_lock);
	282
	283	if (mmap)
	284	rv = mmap(file, vma);
	285
	286	pde_users_dec(pde);
	287	return rv;
	288	}
	289
	290	static int proc_reg_open(struct inode inode, struct file file)
	291	{
	292	struct proc_dir_entry *pde = PDE(inode);
	293	int rv = 0;
	294	int (open)(struct inode , struct file *);
881adb85 AD	295	int (release)(struct inode , struct file *);
	296	struct pde_opener *pdeo;
	297
	298	/*
	299	* What for, you ask? Well, we can have open, rmmod, remove_proc_entry
	300	* sequence. ->release won't be called because ->proc_fops will be
	301	* cleared. Depending on complexity of ->release, consequences vary.
	302	*
	303	* We can't wait for mercy when close will be done for real, it's
	304	* deadlockable: rmmod foo </proc/foo . So, we're going to do ->release
	305	* by hand in remove_proc_entry(). For this, save opener's credentials
	306	* for later.
	307	*/
	308	pdeo = kmalloc(sizeof(struct pde_opener), GFP_KERNEL);
	309	if (!pdeo)
	310	return -ENOMEM;
786d7e16 AD	311
	312	spin_lock(&pde->pde_unload_lock);
	313	if (!pde->proc_fops) {
	314	spin_unlock(&pde->pde_unload_lock);
881adb85	315	kfree(pdeo);
300b994b	316	return -EINVAL;
786d7e16 AD	317	}
	318	pde->pde_users++;
	319	open = pde->proc_fops->open;
881adb85	320	release = pde->proc_fops->release;
786d7e16 AD	321	spin_unlock(&pde->pde_unload_lock);
	322
	323	if (open)
	324	rv = open(inode, file);
	325
881adb85 AD	326	spin_lock(&pde->pde_unload_lock);
	327	if (rv == 0 && release) {
	328	/* To know what to release. */
	329	pdeo->inode = inode;
	330	pdeo->file = file;
	331	/* Strictly for "too late" ->release in proc_reg_release(). */
	332	pdeo->release = release;
	333	list_add(&pdeo->lh, &pde->pde_openers);
	334	} else
	335	kfree(pdeo);
	336	__pde_users_dec(pde);
	337	spin_unlock(&pde->pde_unload_lock);
786d7e16 AD	338	return rv;
	339	}
	340
881adb85 AD	341	static struct pde_opener find_pde_opener(struct proc_dir_entry pde,
	342	struct inode inode, struct file file)
	343	{
	344	struct pde_opener *pdeo;
	345
	346	list_for_each_entry(pdeo, &pde->pde_openers, lh) {
	347	if (pdeo->inode == inode && pdeo->file == file)
	348	return pdeo;
	349	}
	350	return NULL;
	351	}
	352
786d7e16 AD	353	static int proc_reg_release(struct inode inode, struct file file)
	354	{
	355	struct proc_dir_entry *pde = PDE(inode);
	356	int rv = 0;
	357	int (release)(struct inode , struct file *);
881adb85	358	struct pde_opener *pdeo;
786d7e16 AD	359
786d7e16 AD	360	spin_lock(&pde->pde_unload_lock);
881adb85	361	pdeo = find_pde_opener(pde, inode, file);
786d7e16	362	if (!pde->proc_fops) {
881adb85 AD	363	/*
	364	* Can't simply exit, __fput() will think that everything is OK,
	365	* and move on to freeing struct file. remove_proc_entry() will
	366	* find slacker in opener's list and will try to do non-trivial
	367	* things with struct file. Therefore, remove opener from list.
	368	*
	369	* But if opener is removed from list, who will ->release it?
	370	*/
	371	if (pdeo) {
	372	list_del(&pdeo->lh);
	373	spin_unlock(&pde->pde_unload_lock);
	374	rv = pdeo->release(inode, file);
	375	kfree(pdeo);
	376	} else
	377	spin_unlock(&pde->pde_unload_lock);
786d7e16 AD	378	return rv;
	379	}
	380	pde->pde_users++;
	381	release = pde->proc_fops->release;
881adb85 AD	382	if (pdeo) {
	383	list_del(&pdeo->lh);
	384	kfree(pdeo);
	385	}
786d7e16 AD	386	spin_unlock(&pde->pde_unload_lock);
	387
	388	if (release)
	389	rv = release(inode, file);
	390
	391	pde_users_dec(pde);
	392	return rv;
	393	}
	394
	395	static const struct file_operations proc_reg_file_ops = {
	396	.llseek = proc_reg_llseek,
	397	.read = proc_reg_read,
	398	.write = proc_reg_write,
	399	.poll = proc_reg_poll,
	400	.unlocked_ioctl = proc_reg_unlocked_ioctl,
	401	#ifdef CONFIG_COMPAT
	402	.compat_ioctl = proc_reg_compat_ioctl,
	403	#endif
	404	.mmap = proc_reg_mmap,
	405	.open = proc_reg_open,
	406	.release = proc_reg_release,
	407	};
	408
778f3dd5 DM	409	#ifdef CONFIG_COMPAT
	410	static const struct file_operations proc_reg_file_ops_no_compat = {
	411	.llseek = proc_reg_llseek,
	412	.read = proc_reg_read,
	413	.write = proc_reg_write,
	414	.poll = proc_reg_poll,
	415	.unlocked_ioctl = proc_reg_unlocked_ioctl,
	416	.mmap = proc_reg_mmap,
	417	.open = proc_reg_open,
	418	.release = proc_reg_release,
	419	};
	420	#endif
	421
1da177e4 LT	422	struct inode proc_get_inode(struct super_block sb, unsigned int ino,
	423	struct proc_dir_entry *de)
	424	{
	425	struct inode * inode;
	426
a1d4aebb	427	inode = iget_locked(sb, ino);
1da177e4	428	if (!inode)
99b76233	429	return NULL;
a1d4aebb DH	430	if (inode->i_state & I_NEW) {
	431	inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME;
	432	PROC_I(inode)->fd = 0;
	433	PROC_I(inode)->pde = de;
5e971dce AD	434
	435	if (de->mode) {
	436	inode->i_mode = de->mode;
	437	inode->i_uid = de->uid;
	438	inode->i_gid = de->gid;
	439	}
	440	if (de->size)
	441	inode->i_size = de->size;
	442	if (de->nlink)
	443	inode->i_nlink = de->nlink;
	444	if (de->proc_iops)
	445	inode->i_op = de->proc_iops;
	446	if (de->proc_fops) {
	447	if (S_ISREG(inode->i_mode)) {
778f3dd5	448	#ifdef CONFIG_COMPAT
5e971dce AD	449	if (!de->proc_fops->compat_ioctl)
	450	inode->i_fop =
	451	&proc_reg_file_ops_no_compat;
	452	else
778f3dd5	453	#endif
5e971dce AD	454	inode->i_fop = &proc_reg_file_ops;
	455	} else {
	456	inode->i_fop = de->proc_fops;
778f3dd5	457	}
786d7e16	458	}
a1d4aebb	459	unlock_new_inode(inode);
99b76233	460	} else
135d5655	461	pde_put(de);
1da177e4	462	return inode;
1da177e4 LT	463	}
1da177e4 LT	464
07543f5c	465	int proc_fill_super(struct super_block *s)
1da177e4 LT	466	{
	467	struct inode * root_inode;
	468
92d03285	469	s->s_flags \|= MS_NODIRATIME \| MS_NOSUID \| MS_NOEXEC;
1da177e4 LT	470	s->s_blocksize = 1024;
	471	s->s_blocksize_bits = 10;
	472	s->s_magic = PROC_SUPER_MAGIC;
	473	s->s_op = &proc_sops;
	474	s->s_time_gran = 1;
	475
135d5655	476	pde_get(&proc_root);
1da177e4 LT	477	root_inode = proc_get_inode(s, PROC_ROOT_INO, &proc_root);
	478	if (!root_inode)
	479	goto out_no_root;
1da177e4 LT	480	root_inode->i_uid = 0;
	481	root_inode->i_gid = 0;
	482	s->s_root = d_alloc_root(root_inode);
	483	if (!s->s_root)
	484	goto out_no_root;
	485	return 0;
	486
	487	out_no_root:
	488	printk("proc_read_super: get root inode failed\n");
	489	iput(root_inode);
135d5655	490	pde_put(&proc_root);
1da177e4 LT	491	return -ENOMEM;
1da177e4 LT	492	}