Release 4.11 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/fdtable.h>
#include <linux/init.h>
#include <linux/module.h>
#include <linux/fs.h>
#include <linux/security.h>
#include <linux/cred.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>
#include <linux/sysctl.h>
#include <linux/percpu_counter.h>
#include <linux/percpu.h>
#include <linux/hardirq.h>
#include <linux/task_work.h>
#include <linux/ima.h>
#include <linux/swap.h>

#include <linux/atomic.h>

#include "internal.h"

/* sysctl tunables... */

struct files_stat_struct files_stat = {
	.max_files = NR_FILE
};

/* SLAB cache for file structures */

static struct kmem_cache *filp_cachep __read_mostly;


static struct percpu_counter nr_files __cacheline_aligned_in_smp;



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

	put_cred(f->f_cred);
	kmem_cache_free(filp_cachep, f);
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Dipankar Sarma
32
76.19%
1
33.33%
David Howells
7
16.67%
1
33.33%
Eric Dumazet
3
7.14%
1
33.33%
Total
42
100.00%
3
100.00%



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

Contributors
Person
Tokens
Prop
Commits
CommitProp
Andrew Morton
16
53.33%
1
25.00%
Dipankar Sarma
11
36.67%
2
50.00%
Eric Dumazet
3
10.00%
1
25.00%
Total
30
100.00%
4
100.00%

/*
 * Return the total number of open files in the system
 */


static long get_nr_files(void)
{
	return percpu_counter_read_positive(&nr_files);
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Dipankar Sarma
14
93.33%
1
50.00%
Eric Dumazet
1
6.67%
1
50.00%
Total
15
100.00%
2
100.00%

/*
 * Return the maximum number of open files in the system
 */


unsigned long get_max_files(void)
{
	return files_stat.max_files;
}
Contributors
Person
Tokens
Prop
Commits
CommitProp
Dipankar Sarma
11
84.62%
1
50.00%
Eric Dumazet
2
15.38%
1
50.00%
Total
13
100.00%
2
100.00%


EXPORT_SYMBOL_GPL(get_max_files);

/*
 * Handle nr_files sysctl
 */
#if defined(CONFIG_SYSCTL) && defined(CONFIG_PROC_FS)


int proc_nr_files(struct ctl_table *table, int write,
                     void __user *buffer, size_t *lenp, loff_t *ppos)
{
	files_stat.nr_files = get_nr_files();
	return proc_doulongvec_minmax(table, write, buffer, lenp, ppos);
}
Contributors
Person
Tokens
Prop
Commits
CommitProp
Dipankar Sarma
45
95.74%
1
33.33%
Eric Dumazet
1
2.13%
1
33.33%
Joe Perches
1
2.13%
1
33.33%
Total
47
100.00%
3
100.00%

#else


int proc_nr_files(struct ctl_table *table, int write,
                     void __user *buffer, size_t *lenp, loff_t *ppos)
{
	return -ENOSYS;
}
Contributors
Person
Tokens
Prop
Commits
CommitProp
Dipankar Sarma
29
96.67%
1
50.00%
Joe Perches
1
3.33%
1
50.00%
Total
30
100.00%
2
100.00%

#endif

/* Find an unused file structure and return a pointer to it.
 * Returns an error pointer if some error happend e.g. we over file
 * structures limit, run out of memory or operation is not permitted.
 *
 * Be very careful using this.  You are responsible for
 * getting write access to any mount that you might assign
 * to this filp, if it is opened for write.  If this is not
 * done, you will imbalance int the mount's writer count
 * and a warning at __fput() time.
 */


struct file *get_empty_filp(void)
{
	const struct cred *cred = current_cred();
	static long old_max;
	struct file *f;
	int error;

	/*
         * Privileged users can go above max_files
         */
	if (get_nr_files() >= files_stat.max_files && !capable(CAP_SYS_ADMIN)) {
		/*
                 * percpu_counters are inaccurate.  Do an expensive check before
                 * we go and fail.
                 */
		if (percpu_counter_sum_positive(&nr_files) >= files_stat.max_files)
			goto over;
	}

	f = kmem_cache_zalloc(filp_cachep, GFP_KERNEL);
	if (unlikely(!f))
		return ERR_PTR(-ENOMEM);

	percpu_counter_inc(&nr_files);
	f->f_cred = get_cred(cred);
	error = security_file_alloc(f);
	if (unlikely(error)) {
		file_free(f);
		return ERR_PTR(error);
	}

	atomic_long_set(&f->f_count, 1);
	rwlock_init(&f->f_owner.lock);
	spin_lock_init(&f->f_lock);
	mutex_init(&f->f_pos_lock);
	eventpoll_init_file(f);
	/* f->f_version: 0 */
	return f;

over:
	/* Ran out of filps - report that */
	if (get_nr_files() > old_max) {
		pr_info("VFS: file-max limit %lu reached\n", get_max_files());
		old_max = get_nr_files();
	}
	return ERR_PTR(-ENFILE);
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Linus Torvalds (pre-git)
53
25.60%
7
26.92%
Al Viro
41
19.81%
2
7.69%
Dipankar Sarma
28
13.53%
1
3.85%
Andrew Morton
17
8.21%
3
11.54%
Benjamin LaHaise
14
6.76%
1
3.85%
Tetsuo Handa
9
4.35%
1
3.85%
Kirill Korotaev
9
4.35%
1
3.85%
Linus Torvalds
8
3.86%
1
3.85%
Jonathan Corbet
8
3.86%
1
3.85%
David Howells
6
2.90%
1
3.85%
Stephen D. Smalley
4
1.93%
1
3.85%
Eric Dumazet
3
1.45%
1
3.85%
Thomas Gleixner
2
0.97%
1
3.85%
James Morris
2
0.97%
1
3.85%
Greg Kroah-Hartman
1
0.48%
1
3.85%
Cheng Renquan
1
0.48%
1
3.85%
Peter Zijlstra
1
0.48%
1
3.85%
Total
207
100.00%
26
100.00%

/**
 * alloc_file - allocate and initialize a 'struct file'
 *
 * @path: the (dentry, vfsmount) pair for the new file
 * @mode: the mode with which the new file will be opened
 * @fop: the 'struct file_operations' for the new file
 */


struct file *alloc_file(const struct path *path, fmode_t mode,
		const struct file_operations *fop)
{
	struct file *file;

	file = get_empty_filp();
	if (IS_ERR(file))
		return file;

	file->f_path = *path;
	file->f_inode = path->dentry->d_inode;
	file->f_mapping = path->dentry->d_inode->i_mapping;
	if ((mode & FMODE_READ) &&
	     likely(fop->read || fop->read_iter))
		mode |= FMODE_CAN_READ;
	if ((mode & FMODE_WRITE) &&
	     likely(fop->write || fop->write_iter))
		mode |= FMODE_CAN_WRITE;
	file->f_mode = mode;
	file->f_op = fop;
	if ((mode & (FMODE_READ | FMODE_WRITE)) == FMODE_READ)
		i_readcount_inc(path->dentry->d_inode);
	return file;
}
Contributors
Person
Tokens
Prop
Commits
CommitProp
Al Viro
71
45.81%
9
75.00%
Dave Hansen
63
40.65%
1
8.33%
Mimi Zohar
20
12.90%
1
8.33%
Anatol Pomozov
1
0.65%
1
8.33%
Total
155
100.00%
12
100.00%


EXPORT_SYMBOL(alloc_file);

/* the real guts of fput() - releasing the last reference to file
 */


static void __fput(struct file *file)
{
	struct dentry *dentry = file->f_path.dentry;
	struct vfsmount *mnt = file->f_path.mnt;
	struct inode *inode = file->f_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_file(file);

	if (unlikely(file->f_flags & FASYNC)) {
		if (file->f_op->fasync)
			file->f_op->fasync(-1, file, 0);
	}
	ima_file_free(file);
	if (file->f_op->release)
		file->f_op->release(inode, file);
	security_file_free(file);
	if (unlikely(S_ISCHR(inode->i_mode) && inode->i_cdev != NULL &&
		     !(file->f_mode & FMODE_PATH))) {
		cdev_put(inode->i_cdev);
	}
	fops_put(file->f_op);
	put_pid(file->f_owner.pid);
	if ((file->f_mode & (FMODE_READ | FMODE_WRITE)) == FMODE_READ)
		i_readcount_dec(inode);
	if (file->f_mode & FMODE_WRITER) {
		put_write_access(inode);
		__mnt_drop_write(mnt);
	}
	file->f_path.dentry = NULL;
	file->f_path.mnt = NULL;
	file->f_inode = NULL;
	file_free(file);
	dput(dentry);
	mntput(mnt);
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Al Viro
70
27.45%
5
16.67%
Linus Torvalds (pre-git)
58
22.75%
7
23.33%
Mimi Zohar
26
10.20%
2
6.67%
Linus Torvalds
20
7.84%
1
3.33%
Josef 'Jeff' Sipek
12
4.71%
1
3.33%
Miklos Szeredi
11
4.31%
1
3.33%
Dave Hansen
10
3.92%
1
3.33%
Benjamin LaHaise
9
3.53%
1
3.33%
Eric W. Biedermann
9
3.53%
1
3.33%
Theodore Y. Ts'o
7
2.75%
1
3.33%
Davide Libenzi
6
2.35%
2
6.67%
Robert Love
5
1.96%
1
3.33%
Stephen D. Smalley
4
1.57%
1
3.33%
Ingo Molnar
3
1.18%
1
3.33%
David Howells
2
0.78%
1
3.33%
Jeff Layton
1
0.39%
1
3.33%
Greg Kroah-Hartman
1
0.39%
1
3.33%
Andrew Morton
1
0.39%
1
3.33%
Total
255
100.00%
30
100.00%

static LLIST_HEAD(delayed_fput_list);


static void delayed_fput(struct work_struct *unused)
{
	struct llist_node *node = llist_del_all(&delayed_fput_list);
	struct llist_node *next;

	for (; node; node = next) {
		next = llist_next(node);
		__fput(llist_entry(node, struct file, f_u.fu_llist));
	}
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Al Viro
30
50.00%
1
50.00%
Oleg Nesterov
30
50.00%
1
50.00%
Total
60
100.00%
2
100.00%



static void ____fput(struct callback_head *work)
{
	__fput(container_of(work, struct file, f_u.fu_rcuhead));
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Al Viro
26
100.00%
1
100.00%
Total
26
100.00%
1
100.00%

/*
 * If kernel thread really needs to have the final fput() it has done
 * to complete, call this.  The only user right now is the boot - we
 * *do* need to make sure our writes to binaries on initramfs has
 * not left us with opened struct file waiting for __fput() - execve()
 * won't work without that.  Please, don't add more callers without
 * very good reasons; in particular, never call that with locks
 * held and never call that from a thread that might need to do
 * some work on any kind of umount.
 */


void flush_delayed_fput(void)
{
	delayed_fput(NULL);
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Al Viro
12
100.00%
1
100.00%
Total
12
100.00%
1
100.00%

static DECLARE_DELAYED_WORK(delayed_fput_work, delayed_fput);



void fput(struct file *file)
{
	if (atomic_long_dec_and_test(&file->f_count)) {
		struct task_struct *task = current;

		if (likely(!in_interrupt() && !(task->flags & PF_KTHREAD))) {
			init_task_work(&file->f_u.fu_rcuhead, ____fput);
			if (!task_work_add(task, &file->f_u.fu_rcuhead, true))
				return;
			/*
                         * After this task has run exit_task_work(),
                         * task_work_add() will fail.  Fall through to delayed
                         * fput to avoid leaking *file.
                         */
		}

		if (llist_add(&file->f_u.fu_llist, &delayed_fput_list))
			schedule_delayed_work(&delayed_fput_work, 1);
	}
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Al Viro
60
58.25%
3
50.00%
Oleg Nesterov
42
40.78%
2
33.33%
Andrew Morton
1
0.97%
1
16.67%
Total
103
100.00%
6
100.00%

/*
 * synchronous analog of fput(); for kernel threads that might be needed
 * in some umount() (and thus can't use flush_delayed_fput() without
 * risking deadlocks), need to wait for completion of __fput() and know
 * for this specific struct file it won't involve anything that would
 * need them.  Use only if you really need it - at the very least,
 * don't blindly convert fput() by kernel thread to that.
 */


void __fput_sync(struct file *file)
{
	if (atomic_long_dec_and_test(&file->f_count)) {
		struct task_struct *task = current;
		BUG_ON(!(task->flags & PF_KTHREAD));
		__fput(file);
	}
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Al Viro
46
100.00%
2
100.00%
Total
46
100.00%
2
100.00%


EXPORT_SYMBOL(fput);



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

Contributors
Person
Tokens
Prop
Commits
CommitProp
Linus Torvalds (pre-git)
25
78.12%
3
42.86%
Stephen D. Smalley
4
12.50%
1
14.29%
Greg Kroah-Hartman
1
3.12%
1
14.29%
Andrew Morton
1
3.12%
1
14.29%
Al Viro
1
3.12%
1
14.29%
Total
32
100.00%
7
100.00%



void __init files_init(void)
{ 
	filp_cachep = kmem_cache_create("filp", sizeof(struct file), 0,
			SLAB_HWCACHE_ALIGN | SLAB_PANIC, NULL);
	percpu_counter_init(&nr_files, 0, GFP_KERNEL);
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Eric Dumazet
21
53.85%
1
33.33%
Mel Gorman
12
30.77%
1
33.33%
Andi Kleen
6
15.38%
1
33.33%
Total
39
100.00%
3
100.00%

/*
 * One file with associated inode and dcache is very roughly 1K. Per default
 * do not use more than 10% of our memory for files.
 */


void __init files_maxfiles_init(void)
{
	unsigned long n;
	unsigned long memreserve = (totalram_pages - nr_free_pages()) * 3/2;

	memreserve = min(memreserve, totalram_pages - 1);
	n = ((totalram_pages - memreserve) * (PAGE_SIZE / 1024)) / 10;

	files_stat.max_files = max_t(unsigned long, n, NR_FILE);
} 
Contributors
Person
Tokens
Prop
Commits
CommitProp
Mel Gorman
42
60.87%
1
33.33%
Andi Kleen
19
27.54%
1
33.33%
Eric Dumazet
8
11.59%
1
33.33%
Total
69
100.00%
3
100.00%

Overall Contributors
Person
Tokens
Prop
Commits
CommitProp
Al Viro
394
29.56%
17
18.89%
Dipankar Sarma
206
15.45%
2
2.22%
Linus Torvalds (pre-git)
163
12.23%
18
20.00%
Oleg Nesterov
73
5.48%
2
2.22%
Dave Hansen
73
5.48%
2
2.22%
Mel Gorman
58
4.35%
1
1.11%
Eric Dumazet
50
3.75%
3
3.33%
Mimi Zohar
46
3.45%
2
2.22%
Andrew Morton
40
3.00%
4
4.44%
Linus Torvalds
28
2.10%
2
2.22%
Andi Kleen
25
1.88%
1
1.11%
Benjamin LaHaise
23
1.73%
2
2.22%
David Howells
15
1.13%
3
3.33%
Stephen D. Smalley
15
1.13%
1
1.11%
Josef 'Jeff' Sipek
12
0.90%
1
1.11%
Miklos Szeredi
11
0.83%
1
1.11%
Eric W. Biedermann
9
0.68%
1
1.11%
Tetsuo Handa
9
0.68%
1
1.11%
Davide Libenzi
9
0.68%
3
3.33%
Kirill Korotaev
9
0.68%
1
1.11%
Jonathan Corbet
8
0.60%
1
1.11%
Robert Love
8
0.60%
1
1.11%
Theodore Y. Ts'o
7
0.53%
1
1.11%
Ingo Molnar
6
0.45%
2
2.22%
Roland Dreier
5
0.38%
1
1.11%
Dave Jones
4
0.30%
2
2.22%
Eric Paris
3
0.23%
1
1.11%
Randy Dunlap
3
0.23%
1
1.11%
Nicholas Piggin
3
0.23%
1
1.11%
Christoph Hellwig
3
0.23%
1
1.11%
Greg Kroah-Hartman
3
0.23%
1
1.11%
James Morris
2
0.15%
1
1.11%
Joe Perches
2
0.15%
1
1.11%
Thomas Gleixner
2
0.15%
1
1.11%
Arun Sharma
1
0.08%
1
1.11%
Jeff Layton
1
0.08%
1
1.11%
Eric Biggers
1
0.08%
1
1.11%
Peter Zijlstra
1
0.08%
1
1.11%
Cheng Renquan
1
0.08%
1
1.11%
Anatol Pomozov
1
0.08%
1
1.11%
Total
1333
100.00%
90
100.00%
Directory: fs

Information contained on this website is for historical information purposes only and does not indicate or represent copyright ownership.