[net-next-2.6.git] / fs / file_table.c

/*
 *  linux/fs/file_table.c
 *
 *  Copyright (C) 1991, 1992  Linus Torvalds
 *  Copyright (C) 1997 David S. Miller (davem@caip.rutgers.edu)
 */

#include <linux/string.h>
#include <linux/slab.h>
#include <linux/file.h>
#include <linux/init.h>
#include <linux/module.h>
#include <linux/smp_lock.h>
#include <linux/fs.h>
#include <linux/security.h>
#include <linux/eventpoll.h>
#include <linux/rcupdate.h>
#include <linux/mount.h>
#include <linux/capability.h>
#include <linux/cdev.h>
#include <linux/fsnotify.h>

/* sysctl tunables... */
struct files_stat_struct files_stat = {
	.max_files = NR_FILE
};

EXPORT_SYMBOL(files_stat); /* Needed by unix.o */

/* public. Not pretty! */
 __cacheline_aligned_in_smp DEFINE_SPINLOCK(files_lock);

static DEFINE_SPINLOCK(filp_count_lock);

/* slab constructors and destructors are called from arbitrary
 * context and must be fully threaded - use a local spinlock
 * to protect files_stat.nr_files
 */
void filp_ctor(void *objp, struct kmem_cache *cachep, unsigned long cflags)
{
	if ((cflags & (SLAB_CTOR_VERIFY|SLAB_CTOR_CONSTRUCTOR)) ==
	    SLAB_CTOR_CONSTRUCTOR) {
		unsigned long flags;
		spin_lock_irqsave(&filp_count_lock, flags);
		files_stat.nr_files++;
		spin_unlock_irqrestore(&filp_count_lock, flags);
	}
}

void filp_dtor(void *objp, struct kmem_cache *cachep, unsigned long dflags)
{
	unsigned long flags;
	spin_lock_irqsave(&filp_count_lock, flags);
	files_stat.nr_files--;
	spin_unlock_irqrestore(&filp_count_lock, flags);
}

static inline void file_free_rcu(struct rcu_head *head)
{
	struct file *f =  container_of(head, struct file, f_u.fu_rcuhead);
	kmem_cache_free(filp_cachep, f);
}

static inline void file_free(struct file *f)
{
	call_rcu(&f->f_u.fu_rcuhead, file_free_rcu);
}

/* Find an unused file structure and return a pointer to it.
 * Returns NULL, if there are no more free file structures or
 * we run out of memory.
 */
struct file *get_empty_filp(void)
{
	static int old_max;
	struct file * f;

	/*
	 * Privileged users can go above max_files
	 */
	if (files_stat.nr_files >= files_stat.max_files &&
				!capable(CAP_SYS_ADMIN))
		goto over;

	f = kmem_cache_alloc(filp_cachep, GFP_KERNEL);
	if (f == NULL)
		goto fail;

	memset(f, 0, sizeof(*f));
	if (security_file_alloc(f))
		goto fail_sec;

	eventpoll_init_file(f);
	atomic_set(&f->f_count, 1);
	f->f_uid = current->fsuid;
	f->f_gid = current->fsgid;
	rwlock_init(&f->f_owner.lock);
	/* f->f_version: 0 */
	INIT_LIST_HEAD(&f->f_u.fu_list);
	return f;

over:
	/* Ran out of filps - report that */
	if (files_stat.nr_files > old_max) {
		printk(KERN_INFO "VFS: file-max limit %d reached\n",
					files_stat.max_files);
		old_max = files_stat.nr_files;
	}
	goto fail;

fail_sec:
	file_free(f);
fail:
	return NULL;
}

EXPORT_SYMBOL(get_empty_filp);

void fastcall fput(struct file *file)
{
	if (atomic_dec_and_test(&file->f_count))
		__fput(file);
}

EXPORT_SYMBOL(fput);

/* __fput is called from task context when aio completion releases the last
 * last use of a struct file *.  Do not use otherwise.
 */
void fastcall __fput(struct file *file)
{
	struct dentry *dentry = file->f_dentry;
	struct vfsmount *mnt = file->f_vfsmnt;
	struct inode *inode = dentry->d_inode;

	might_sleep();

	fsnotify_close(file);
	/*
	 * The function eventpoll_release() should be the first called
	 * in the file cleanup chain.
	 */
	eventpoll_release(file);
	locks_remove_flock(file);

	if (file->f_op && file->f_op->release)
		file->f_op->release(inode, file);
	security_file_free(file);
	if (unlikely(inode->i_cdev != NULL))
		cdev_put(inode->i_cdev);
	fops_put(file->f_op);
	if (file->f_mode & FMODE_WRITE)
		put_write_access(inode);
	file_kill(file);
	file->f_dentry = NULL;
	file->f_vfsmnt = NULL;
	file_free(file);
	dput(dentry);
	mntput(mnt);
}

struct file fastcall *fget(unsigned int fd)
{
	struct file *file;
	struct files_struct *files = current->files;

	rcu_read_lock();
	file = fcheck_files(files, fd);
	if (file) {
		if (!atomic_inc_not_zero(&file->f_count)) {
			/* File object ref couldn't be taken */
			rcu_read_unlock();
			return NULL;
		}
	}
	rcu_read_unlock();

	return file;
}

EXPORT_SYMBOL(fget);

/*
 * Lightweight file lookup - no refcnt increment if fd table isn't shared. 
 * You can use this only if it is guranteed that the current task already 
 * holds a refcnt to that file. That check has to be done at fget() only
 * and a flag is returned to be passed to the corresponding fput_light().
 * There must not be a cloning between an fget_light/fput_light pair.
 */
struct file fastcall *fget_light(unsigned int fd, int *fput_needed)
{
	struct file *file;
	struct files_struct *files = current->files;

	*fput_needed = 0;
	if (likely((atomic_read(&files->count) == 1))) {
		file = fcheck_files(files, fd);
	} else {
		rcu_read_lock();
		file = fcheck_files(files, fd);
		if (file) {
			if (atomic_inc_not_zero(&file->f_count))
				*fput_needed = 1;
			else
				/* Didn't get the reference, someone's freed */
				file = NULL;
		}
		rcu_read_unlock();
	}

	return file;
}


void put_filp(struct file *file)
{
	if (atomic_dec_and_test(&file->f_count)) {
		security_file_free(file);
		file_kill(file);
		file_free(file);
	}
}

void file_move(struct file *file, struct list_head *list)
{
	if (!list)
		return;
	file_list_lock();
	list_move(&file->f_u.fu_list, list);
	file_list_unlock();
}

void file_kill(struct file *file)
{
	if (!list_empty(&file->f_u.fu_list)) {
		file_list_lock();
		list_del_init(&file->f_u.fu_list);
		file_list_unlock();
	}
}

int fs_may_remount_ro(struct super_block *sb)
{
	struct list_head *p;

	/* Check that no files are currently opened for writing. */
	file_list_lock();
	list_for_each(p, &sb->s_files) {
		struct file *file = list_entry(p, struct file, f_u.fu_list);
		struct inode *inode = file->f_dentry->d_inode;

		/* File with pending delete? */
		if (inode->i_nlink == 0)
			goto too_bad;

		/* Writeable file? */
		if (S_ISREG(inode->i_mode) && (file->f_mode & FMODE_WRITE))
			goto too_bad;
	}
	file_list_unlock();
	return 1; /* Tis' cool bro. */
too_bad:
	file_list_unlock();
	return 0;
}

void __init files_init(unsigned long mempages)
{ 
	int n; 
	/* One file with associated inode and dcache is very roughly 1K. 
	 * Per default don't use more than 10% of our memory for files. 
	 */ 

	n = (mempages * (PAGE_SIZE / 1024)) / 10;
	files_stat.max_files = n; 
	if (files_stat.max_files < NR_FILE)
		files_stat.max_files = NR_FILE;
	files_defer_init();
}
Commit	Line	Data
1da177e4 LT	1	/*
	2	* linux/fs/file_table.c
	3	*
	4	* Copyright (C) 1991, 1992 Linus Torvalds
	5	* Copyright (C) 1997 David S. Miller (davem@caip.rutgers.edu)
	6	*/
	7
	8	#include <linux/string.h>
	9	#include <linux/slab.h>
	10	#include <linux/file.h>
	11	#include <linux/init.h>
	12	#include <linux/module.h>
	13	#include <linux/smp_lock.h>
	14	#include <linux/fs.h>
	15	#include <linux/security.h>
	16	#include <linux/eventpoll.h>
ab2af1f5	17	#include <linux/rcupdate.h>
1da177e4	18	#include <linux/mount.h>
16f7e0fe	19	#include <linux/capability.h>
1da177e4	20	#include <linux/cdev.h>
0eeca283	21	#include <linux/fsnotify.h>
1da177e4 LT	22
	23	/* sysctl tunables... */
	24	struct files_stat_struct files_stat = {
	25	.max_files = NR_FILE
	26	};
	27
	28	EXPORT_SYMBOL(files_stat); /* Needed by unix.o */
	29
	30	/* public. Not pretty! */
	31	__cacheline_aligned_in_smp DEFINE_SPINLOCK(files_lock);
	32
	33	static DEFINE_SPINLOCK(filp_count_lock);
	34
	35	/* slab constructors and destructors are called from arbitrary
	36	* context and must be fully threaded - use a local spinlock
	37	* to protect files_stat.nr_files
	38	*/
2109a2d1	39	void filp_ctor(void objp, struct kmem_cache cachep, unsigned long cflags)
1da177e4 LT	40	{
	41	if ((cflags & (SLAB_CTOR_VERIFY\|SLAB_CTOR_CONSTRUCTOR)) ==
	42	SLAB_CTOR_CONSTRUCTOR) {
	43	unsigned long flags;
	44	spin_lock_irqsave(&filp_count_lock, flags);
	45	files_stat.nr_files++;
	46	spin_unlock_irqrestore(&filp_count_lock, flags);
	47	}
	48	}
	49
2109a2d1	50	void filp_dtor(void objp, struct kmem_cache cachep, unsigned long dflags)
1da177e4 LT	51	{
	52	unsigned long flags;
	53	spin_lock_irqsave(&filp_count_lock, flags);
	54	files_stat.nr_files--;
	55	spin_unlock_irqrestore(&filp_count_lock, flags);
	56	}
	57
ab2af1f5	58	static inline void file_free_rcu(struct rcu_head *head)
1da177e4	59	{
2f512016	60	struct file *f = container_of(head, struct file, f_u.fu_rcuhead);
1da177e4 LT	61	kmem_cache_free(filp_cachep, f);
	62	}
	63
ab2af1f5 DS	64	static inline void file_free(struct file *f)
ab2af1f5 DS	65	{
2f512016	66	call_rcu(&f->f_u.fu_rcuhead, file_free_rcu);
ab2af1f5 DS	67	}
ab2af1f5 DS	68
1da177e4 LT	69	/* Find an unused file structure and return a pointer to it.
	70	* Returns NULL, if there are no more free file structures or
	71	* we run out of memory.
	72	*/
	73	struct file *get_empty_filp(void)
	74	{
af4d2ecb	75	static int old_max;
1da177e4 LT	76	struct file * f;
	77
	78	/*
	79	* Privileged users can go above max_files
	80	*/
af4d2ecb KK	81	if (files_stat.nr_files >= files_stat.max_files &&
	82	!capable(CAP_SYS_ADMIN))
	83	goto over;
	84
	85	f = kmem_cache_alloc(filp_cachep, GFP_KERNEL);
	86	if (f == NULL)
	87	goto fail;
	88
	89	memset(f, 0, sizeof(*f));
	90	if (security_file_alloc(f))
	91	goto fail_sec;
1da177e4	92
af4d2ecb KK	93	eventpoll_init_file(f);
	94	atomic_set(&f->f_count, 1);
	95	f->f_uid = current->fsuid;
	96	f->f_gid = current->fsgid;
	97	rwlock_init(&f->f_owner.lock);
	98	/* f->f_version: 0 */
2f512016	99	INIT_LIST_HEAD(&f->f_u.fu_list);
af4d2ecb KK	100	return f;
	101
	102	over:
1da177e4	103	/* Ran out of filps - report that */
af4d2ecb	104	if (files_stat.nr_files > old_max) {
1da177e4 LT	105	printk(KERN_INFO "VFS: file-max limit %d reached\n",
1da177e4 LT	106	files_stat.max_files);
af4d2ecb	107	old_max = files_stat.nr_files;
1da177e4	108	}
af4d2ecb KK	109	goto fail;
	110
	111	fail_sec:
	112	file_free(f);
1da177e4 LT	113	fail:
	114	return NULL;
	115	}
	116
	117	EXPORT_SYMBOL(get_empty_filp);
	118
	119	void fastcall fput(struct file *file)
	120	{
095975da	121	if (atomic_dec_and_test(&file->f_count))
1da177e4 LT	122	__fput(file);
	123	}
	124
	125	EXPORT_SYMBOL(fput);
	126
	127	/* __fput is called from task context when aio completion releases the last
	128	* last use of a struct file *. Do not use otherwise.
	129	*/
	130	void fastcall __fput(struct file *file)
	131	{
	132	struct dentry *dentry = file->f_dentry;
	133	struct vfsmount *mnt = file->f_vfsmnt;
	134	struct inode *inode = dentry->d_inode;
	135
	136	might_sleep();
0eeca283 RL	137
0eeca283 RL	138	fsnotify_close(file);
1da177e4 LT	139	/*
	140	* The function eventpoll_release() should be the first called
	141	* in the file cleanup chain.
	142	*/
	143	eventpoll_release(file);
	144	locks_remove_flock(file);
	145
	146	if (file->f_op && file->f_op->release)
	147	file->f_op->release(inode, file);
	148	security_file_free(file);
	149	if (unlikely(inode->i_cdev != NULL))
	150	cdev_put(inode->i_cdev);
	151	fops_put(file->f_op);
	152	if (file->f_mode & FMODE_WRITE)
	153	put_write_access(inode);
	154	file_kill(file);
	155	file->f_dentry = NULL;
	156	file->f_vfsmnt = NULL;
	157	file_free(file);
	158	dput(dentry);
	159	mntput(mnt);
	160	}
	161
	162	struct file fastcall *fget(unsigned int fd)
	163	{
	164	struct file *file;
	165	struct files_struct *files = current->files;
	166
ab2af1f5	167	rcu_read_lock();
1da177e4	168	file = fcheck_files(files, fd);
ab2af1f5	169	if (file) {
095975da	170	if (!atomic_inc_not_zero(&file->f_count)) {
ab2af1f5 DS	171	/* File object ref couldn't be taken */
	172	rcu_read_unlock();
	173	return NULL;
	174	}
	175	}
	176	rcu_read_unlock();
	177
1da177e4 LT	178	return file;
	179	}
	180
	181	EXPORT_SYMBOL(fget);
	182
	183	/*
	184	* Lightweight file lookup - no refcnt increment if fd table isn't shared.
	185	* You can use this only if it is guranteed that the current task already
	186	* holds a refcnt to that file. That check has to be done at fget() only
	187	* and a flag is returned to be passed to the corresponding fput_light().
	188	* There must not be a cloning between an fget_light/fput_light pair.
	189	*/
	190	struct file fastcall fget_light(unsigned int fd, int fput_needed)
	191	{
	192	struct file *file;
	193	struct files_struct *files = current->files;
	194
	195	*fput_needed = 0;
	196	if (likely((atomic_read(&files->count) == 1))) {
	197	file = fcheck_files(files, fd);
	198	} else {
ab2af1f5	199	rcu_read_lock();
1da177e4 LT	200	file = fcheck_files(files, fd);
1da177e4 LT	201	if (file) {
095975da	202	if (atomic_inc_not_zero(&file->f_count))
ab2af1f5 DS	203	*fput_needed = 1;
	204	else
	205	/* Didn't get the reference, someone's freed */
	206	file = NULL;
1da177e4	207	}
ab2af1f5	208	rcu_read_unlock();
1da177e4	209	}
ab2af1f5	210
1da177e4 LT	211	return file;
	212	}
	213
	214
	215	void put_filp(struct file *file)
	216	{
095975da	217	if (atomic_dec_and_test(&file->f_count)) {
1da177e4 LT	218	security_file_free(file);
	219	file_kill(file);
	220	file_free(file);
	221	}
	222	}
	223
	224	void file_move(struct file file, struct list_head list)
	225	{
	226	if (!list)
	227	return;
	228	file_list_lock();
2f512016	229	list_move(&file->f_u.fu_list, list);
1da177e4 LT	230	file_list_unlock();
	231	}
	232
	233	void file_kill(struct file *file)
	234	{
2f512016	235	if (!list_empty(&file->f_u.fu_list)) {
1da177e4	236	file_list_lock();
2f512016	237	list_del_init(&file->f_u.fu_list);
1da177e4 LT	238	file_list_unlock();
	239	}
	240	}
	241
	242	int fs_may_remount_ro(struct super_block *sb)
	243	{
	244	struct list_head *p;
	245
	246	/* Check that no files are currently opened for writing. */
	247	file_list_lock();
	248	list_for_each(p, &sb->s_files) {
2f512016	249	struct file *file = list_entry(p, struct file, f_u.fu_list);
1da177e4 LT	250	struct inode *inode = file->f_dentry->d_inode;
	251
	252	/* File with pending delete? */
	253	if (inode->i_nlink == 0)
	254	goto too_bad;
	255
	256	/* Writeable file? */
	257	if (S_ISREG(inode->i_mode) && (file->f_mode & FMODE_WRITE))
	258	goto too_bad;
	259	}
	260	file_list_unlock();
	261	return 1; /* Tis' cool bro. */
	262	too_bad:
	263	file_list_unlock();
	264	return 0;
	265	}
	266
	267	void __init files_init(unsigned long mempages)
	268	{
	269	int n;
	270	/* One file with associated inode and dcache is very roughly 1K.
	271	* Per default don't use more than 10% of our memory for files.
	272	*/
	273
	274	n = (mempages * (PAGE_SIZE / 1024)) / 10;
	275	files_stat.max_files = n;
	276	if (files_stat.max_files < NR_FILE)
	277	files_stat.max_files = NR_FILE;
ab2af1f5	278	files_defer_init();
1da177e4	279	}