2005-04-16 16:20:36 -06:00
|
|
|
/*
|
|
|
|
* linux/fs/locks.c
|
|
|
|
*
|
|
|
|
* Provide support for fcntl()'s F_GETLK, F_SETLK, and F_SETLKW calls.
|
|
|
|
* Doug Evans (dje@spiff.uucp), August 07, 1992
|
|
|
|
*
|
|
|
|
* Deadlock detection added.
|
|
|
|
* FIXME: one thing isn't handled yet:
|
|
|
|
* - mandatory locks (requires lots of changes elsewhere)
|
|
|
|
* Kelly Carmichael (kelly@[142.24.8.65]), September 17, 1994.
|
|
|
|
*
|
|
|
|
* Miscellaneous edits, and a total rewrite of posix_lock_file() code.
|
|
|
|
* Kai Petzke (wpp@marie.physik.tu-berlin.de), 1994
|
|
|
|
*
|
|
|
|
* Converted file_lock_table to a linked list from an array, which eliminates
|
|
|
|
* the limits on how many active file locks are open.
|
|
|
|
* Chad Page (pageone@netcom.com), November 27, 1994
|
|
|
|
*
|
|
|
|
* Removed dependency on file descriptors. dup()'ed file descriptors now
|
|
|
|
* get the same locks as the original file descriptors, and a close() on
|
|
|
|
* any file descriptor removes ALL the locks on the file for the current
|
|
|
|
* process. Since locks still depend on the process id, locks are inherited
|
|
|
|
* after an exec() but not after a fork(). This agrees with POSIX, and both
|
|
|
|
* BSD and SVR4 practice.
|
|
|
|
* Andy Walker (andy@lysaker.kvaerner.no), February 14, 1995
|
|
|
|
*
|
|
|
|
* Scrapped free list which is redundant now that we allocate locks
|
|
|
|
* dynamically with kmalloc()/kfree().
|
|
|
|
* Andy Walker (andy@lysaker.kvaerner.no), February 21, 1995
|
|
|
|
*
|
|
|
|
* Implemented two lock personalities - FL_FLOCK and FL_POSIX.
|
|
|
|
*
|
|
|
|
* FL_POSIX locks are created with calls to fcntl() and lockf() through the
|
|
|
|
* fcntl() system call. They have the semantics described above.
|
|
|
|
*
|
|
|
|
* FL_FLOCK locks are created with calls to flock(), through the flock()
|
|
|
|
* system call, which is new. Old C libraries implement flock() via fcntl()
|
|
|
|
* and will continue to use the old, broken implementation.
|
|
|
|
*
|
|
|
|
* FL_FLOCK locks follow the 4.4 BSD flock() semantics. They are associated
|
|
|
|
* with a file pointer (filp). As a result they can be shared by a parent
|
|
|
|
* process and its children after a fork(). They are removed when the last
|
|
|
|
* file descriptor referring to the file pointer is closed (unless explicitly
|
|
|
|
* unlocked).
|
|
|
|
*
|
|
|
|
* FL_FLOCK locks never deadlock, an existing lock is always removed before
|
|
|
|
* upgrading from shared to exclusive (or vice versa). When this happens
|
|
|
|
* any processes blocked by the current lock are woken up and allowed to
|
|
|
|
* run before the new lock is applied.
|
|
|
|
* Andy Walker (andy@lysaker.kvaerner.no), June 09, 1995
|
|
|
|
*
|
|
|
|
* Removed some race conditions in flock_lock_file(), marked other possible
|
|
|
|
* races. Just grep for FIXME to see them.
|
|
|
|
* Dmitry Gorodchanin (pgmdsg@ibi.com), February 09, 1996.
|
|
|
|
*
|
|
|
|
* Addressed Dmitry's concerns. Deadlock checking no longer recursive.
|
|
|
|
* Lock allocation changed to GFP_ATOMIC as we can't afford to sleep
|
|
|
|
* once we've checked for blocking and deadlocking.
|
|
|
|
* Andy Walker (andy@lysaker.kvaerner.no), April 03, 1996.
|
|
|
|
*
|
|
|
|
* Initial implementation of mandatory locks. SunOS turned out to be
|
|
|
|
* a rotten model, so I implemented the "obvious" semantics.
|
2011-08-14 18:02:26 -06:00
|
|
|
* See 'Documentation/filesystems/mandatory-locking.txt' for details.
|
2005-04-16 16:20:36 -06:00
|
|
|
* Andy Walker (andy@lysaker.kvaerner.no), April 06, 1996.
|
|
|
|
*
|
|
|
|
* Don't allow mandatory locks on mmap()'ed files. Added simple functions to
|
|
|
|
* check if a file has mandatory locks, used by mmap(), open() and creat() to
|
|
|
|
* see if system call should be rejected. Ref. HP-UX/SunOS/Solaris Reference
|
|
|
|
* Manual, Section 2.
|
|
|
|
* Andy Walker (andy@lysaker.kvaerner.no), April 09, 1996.
|
|
|
|
*
|
|
|
|
* Tidied up block list handling. Added '/proc/locks' interface.
|
|
|
|
* Andy Walker (andy@lysaker.kvaerner.no), April 24, 1996.
|
|
|
|
*
|
|
|
|
* Fixed deadlock condition for pathological code that mixes calls to
|
|
|
|
* flock() and fcntl().
|
|
|
|
* Andy Walker (andy@lysaker.kvaerner.no), April 29, 1996.
|
|
|
|
*
|
|
|
|
* Allow only one type of locking scheme (FL_POSIX or FL_FLOCK) to be in use
|
|
|
|
* for a given file at a time. Changed the CONFIG_LOCK_MANDATORY scheme to
|
|
|
|
* guarantee sensible behaviour in the case where file system modules might
|
|
|
|
* be compiled with different options than the kernel itself.
|
|
|
|
* Andy Walker (andy@lysaker.kvaerner.no), May 15, 1996.
|
|
|
|
*
|
|
|
|
* Added a couple of missing wake_up() calls. Thanks to Thomas Meckel
|
|
|
|
* (Thomas.Meckel@mni.fh-giessen.de) for spotting this.
|
|
|
|
* Andy Walker (andy@lysaker.kvaerner.no), May 15, 1996.
|
|
|
|
*
|
|
|
|
* Changed FL_POSIX locks to use the block list in the same way as FL_FLOCK
|
|
|
|
* locks. Changed process synchronisation to avoid dereferencing locks that
|
|
|
|
* have already been freed.
|
|
|
|
* Andy Walker (andy@lysaker.kvaerner.no), Sep 21, 1996.
|
|
|
|
*
|
|
|
|
* Made the block list a circular list to minimise searching in the list.
|
|
|
|
* Andy Walker (andy@lysaker.kvaerner.no), Sep 25, 1996.
|
|
|
|
*
|
|
|
|
* Made mandatory locking a mount option. Default is not to allow mandatory
|
|
|
|
* locking.
|
|
|
|
* Andy Walker (andy@lysaker.kvaerner.no), Oct 04, 1996.
|
|
|
|
*
|
|
|
|
* Some adaptations for NFS support.
|
|
|
|
* Olaf Kirch (okir@monad.swb.de), Dec 1996,
|
|
|
|
*
|
|
|
|
* Fixed /proc/locks interface so that we can't overrun the buffer we are handed.
|
|
|
|
* Andy Walker (andy@lysaker.kvaerner.no), May 12, 1997.
|
|
|
|
*
|
|
|
|
* Use slab allocator instead of kmalloc/kfree.
|
|
|
|
* Use generic list implementation from <linux/list.h>.
|
|
|
|
* Sped up posix_locks_deadlock by only considering blocked locks.
|
|
|
|
* Matthew Wilcox <willy@debian.org>, March, 2000.
|
|
|
|
*
|
|
|
|
* Leases and LOCK_MAND
|
|
|
|
* Matthew Wilcox <willy@debian.org>, June, 2000.
|
|
|
|
* Stephen Rothwell <sfr@canb.auug.org.au>, June, 2000.
|
|
|
|
*/
|
|
|
|
|
|
|
|
#include <linux/capability.h>
|
|
|
|
#include <linux/file.h>
|
2008-04-24 05:44:08 -06:00
|
|
|
#include <linux/fdtable.h>
|
2005-04-16 16:20:36 -06:00
|
|
|
#include <linux/fs.h>
|
|
|
|
#include <linux/init.h>
|
|
|
|
#include <linux/module.h>
|
|
|
|
#include <linux/security.h>
|
|
|
|
#include <linux/slab.h>
|
|
|
|
#include <linux/syscalls.h>
|
|
|
|
#include <linux/time.h>
|
2005-09-16 20:28:13 -06:00
|
|
|
#include <linux/rcupdate.h>
|
2008-01-16 17:07:08 -07:00
|
|
|
#include <linux/pid_namespace.h>
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
#include <asm/uaccess.h>
|
|
|
|
|
|
|
|
#define IS_POSIX(fl) (fl->fl_flags & FL_POSIX)
|
|
|
|
#define IS_FLOCK(fl) (fl->fl_flags & FL_FLOCK)
|
|
|
|
#define IS_LEASE(fl) (fl->fl_flags & FL_LEASE)
|
|
|
|
|
2011-07-26 18:10:51 -06:00
|
|
|
static bool lease_breaking(struct file_lock *fl)
|
|
|
|
{
|
2011-07-26 16:25:49 -06:00
|
|
|
return fl->fl_flags & (FL_UNLOCK_PENDING | FL_DOWNGRADE_PENDING);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int target_leasetype(struct file_lock *fl)
|
|
|
|
{
|
|
|
|
if (fl->fl_flags & FL_UNLOCK_PENDING)
|
|
|
|
return F_UNLCK;
|
|
|
|
if (fl->fl_flags & FL_DOWNGRADE_PENDING)
|
|
|
|
return F_RDLCK;
|
|
|
|
return fl->fl_type;
|
2011-07-26 18:10:51 -06:00
|
|
|
}
|
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
int leases_enable = 1;
|
|
|
|
int lease_break_time = 45;
|
|
|
|
|
|
|
|
#define for_each_lock(inode, lockp) \
|
|
|
|
for (lockp = &inode->i_flock; *lockp != NULL; lockp = &(*lockp)->fl_next)
|
|
|
|
|
2006-03-20 11:44:40 -07:00
|
|
|
static LIST_HEAD(file_lock_list);
|
2005-04-16 16:20:36 -06:00
|
|
|
static LIST_HEAD(blocked_list);
|
2010-10-27 13:39:58 -06:00
|
|
|
static DEFINE_SPINLOCK(file_lock_lock);
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2010-09-18 07:09:31 -06:00
|
|
|
/*
|
|
|
|
* Protects the two list heads above, plus the inode->i_flock list
|
|
|
|
*/
|
|
|
|
void lock_flocks(void)
|
|
|
|
{
|
2010-10-27 13:39:58 -06:00
|
|
|
spin_lock(&file_lock_lock);
|
2010-09-18 07:09:31 -06:00
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(lock_flocks);
|
|
|
|
|
|
|
|
void unlock_flocks(void)
|
|
|
|
{
|
2010-10-27 13:39:58 -06:00
|
|
|
spin_unlock(&file_lock_lock);
|
2010-09-18 07:09:31 -06:00
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(unlock_flocks);
|
|
|
|
|
2006-12-06 21:33:20 -07:00
|
|
|
static struct kmem_cache *filelock_cache __read_mostly;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2011-07-07 05:06:09 -06:00
|
|
|
static void locks_init_lock_heads(struct file_lock *fl)
|
2011-07-06 04:33:55 -06:00
|
|
|
{
|
2011-07-07 05:06:09 -06:00
|
|
|
INIT_LIST_HEAD(&fl->fl_link);
|
|
|
|
INIT_LIST_HEAD(&fl->fl_block);
|
|
|
|
init_waitqueue_head(&fl->fl_wait);
|
2011-07-06 04:33:55 -06:00
|
|
|
}
|
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
/* Allocate an empty lock structure. */
|
2010-10-27 07:46:08 -06:00
|
|
|
struct file_lock *locks_alloc_lock(void)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
2011-07-07 05:06:09 -06:00
|
|
|
struct file_lock *fl = kmem_cache_zalloc(filelock_cache, GFP_KERNEL);
|
2011-07-06 04:33:55 -06:00
|
|
|
|
|
|
|
if (fl)
|
2011-07-07 05:06:09 -06:00
|
|
|
locks_init_lock_heads(fl);
|
2011-07-06 04:33:55 -06:00
|
|
|
|
|
|
|
return fl;
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
2010-10-27 07:46:08 -06:00
|
|
|
EXPORT_SYMBOL_GPL(locks_alloc_lock);
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2009-03-31 14:12:56 -06:00
|
|
|
void locks_release_private(struct file_lock *fl)
|
2006-03-20 11:44:05 -07:00
|
|
|
{
|
|
|
|
if (fl->fl_ops) {
|
|
|
|
if (fl->fl_ops->fl_release_private)
|
|
|
|
fl->fl_ops->fl_release_private(fl);
|
|
|
|
fl->fl_ops = NULL;
|
|
|
|
}
|
2012-08-01 05:56:16 -06:00
|
|
|
fl->fl_lmops = NULL;
|
2006-03-20 11:44:05 -07:00
|
|
|
|
|
|
|
}
|
2009-03-31 14:12:56 -06:00
|
|
|
EXPORT_SYMBOL_GPL(locks_release_private);
|
2006-03-20 11:44:05 -07:00
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
/* Free a lock which is not in use. */
|
2010-10-30 15:31:15 -06:00
|
|
|
void locks_free_lock(struct file_lock *fl)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
2006-03-31 03:30:29 -07:00
|
|
|
BUG_ON(waitqueue_active(&fl->fl_wait));
|
|
|
|
BUG_ON(!list_empty(&fl->fl_block));
|
|
|
|
BUG_ON(!list_empty(&fl->fl_link));
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2006-03-20 11:44:05 -07:00
|
|
|
locks_release_private(fl);
|
2005-04-16 16:20:36 -06:00
|
|
|
kmem_cache_free(filelock_cache, fl);
|
|
|
|
}
|
2010-10-30 15:31:15 -06:00
|
|
|
EXPORT_SYMBOL(locks_free_lock);
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
void locks_init_lock(struct file_lock *fl)
|
|
|
|
{
|
2011-07-07 05:06:09 -06:00
|
|
|
memset(fl, 0, sizeof(struct file_lock));
|
|
|
|
locks_init_lock_heads(fl);
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
|
|
|
|
EXPORT_SYMBOL(locks_init_lock);
|
|
|
|
|
2006-03-20 11:44:05 -07:00
|
|
|
static void locks_copy_private(struct file_lock *new, struct file_lock *fl)
|
|
|
|
{
|
|
|
|
if (fl->fl_ops) {
|
|
|
|
if (fl->fl_ops->fl_copy_lock)
|
|
|
|
fl->fl_ops->fl_copy_lock(new, fl);
|
|
|
|
new->fl_ops = fl->fl_ops;
|
|
|
|
}
|
2010-10-31 06:35:31 -06:00
|
|
|
if (fl->fl_lmops)
|
2006-03-20 11:44:05 -07:00
|
|
|
new->fl_lmops = fl->fl_lmops;
|
|
|
|
}
|
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
/*
|
|
|
|
* Initialize a new lock from an existing file_lock structure.
|
|
|
|
*/
|
2008-04-24 08:08:22 -06:00
|
|
|
void __locks_copy_lock(struct file_lock *new, const struct file_lock *fl)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
|
|
|
new->fl_owner = fl->fl_owner;
|
|
|
|
new->fl_pid = fl->fl_pid;
|
2006-03-20 11:44:38 -07:00
|
|
|
new->fl_file = NULL;
|
2005-04-16 16:20:36 -06:00
|
|
|
new->fl_flags = fl->fl_flags;
|
|
|
|
new->fl_type = fl->fl_type;
|
|
|
|
new->fl_start = fl->fl_start;
|
|
|
|
new->fl_end = fl->fl_end;
|
2006-03-20 11:44:38 -07:00
|
|
|
new->fl_ops = NULL;
|
|
|
|
new->fl_lmops = NULL;
|
|
|
|
}
|
2008-04-25 16:32:51 -06:00
|
|
|
EXPORT_SYMBOL(__locks_copy_lock);
|
2006-03-20 11:44:38 -07:00
|
|
|
|
|
|
|
void locks_copy_lock(struct file_lock *new, struct file_lock *fl)
|
|
|
|
{
|
|
|
|
locks_release_private(new);
|
|
|
|
|
|
|
|
__locks_copy_lock(new, fl);
|
|
|
|
new->fl_file = fl->fl_file;
|
2005-04-16 16:20:36 -06:00
|
|
|
new->fl_ops = fl->fl_ops;
|
|
|
|
new->fl_lmops = fl->fl_lmops;
|
2006-03-20 11:44:05 -07:00
|
|
|
|
|
|
|
locks_copy_private(new, fl);
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
|
|
|
|
EXPORT_SYMBOL(locks_copy_lock);
|
|
|
|
|
|
|
|
static inline int flock_translate_cmd(int cmd) {
|
|
|
|
if (cmd & LOCK_MAND)
|
|
|
|
return cmd & (LOCK_MAND | LOCK_RW);
|
|
|
|
switch (cmd) {
|
|
|
|
case LOCK_SH:
|
|
|
|
return F_RDLCK;
|
|
|
|
case LOCK_EX:
|
|
|
|
return F_WRLCK;
|
|
|
|
case LOCK_UN:
|
|
|
|
return F_UNLCK;
|
|
|
|
}
|
|
|
|
return -EINVAL;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Fill in a file_lock structure with an appropriate FLOCK lock. */
|
|
|
|
static int flock_make_lock(struct file *filp, struct file_lock **lock,
|
|
|
|
unsigned int cmd)
|
|
|
|
{
|
|
|
|
struct file_lock *fl;
|
|
|
|
int type = flock_translate_cmd(cmd);
|
|
|
|
if (type < 0)
|
|
|
|
return type;
|
|
|
|
|
|
|
|
fl = locks_alloc_lock();
|
|
|
|
if (fl == NULL)
|
|
|
|
return -ENOMEM;
|
|
|
|
|
|
|
|
fl->fl_file = filp;
|
|
|
|
fl->fl_pid = current->tgid;
|
|
|
|
fl->fl_flags = FL_FLOCK;
|
|
|
|
fl->fl_type = type;
|
|
|
|
fl->fl_end = OFFSET_MAX;
|
|
|
|
|
|
|
|
*lock = fl;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2012-07-23 13:17:17 -06:00
|
|
|
static int assign_type(struct file_lock *fl, long type)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
|
|
|
switch (type) {
|
|
|
|
case F_RDLCK:
|
|
|
|
case F_WRLCK:
|
|
|
|
case F_UNLCK:
|
|
|
|
fl->fl_type = type;
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
return -EINVAL;
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Verify a "struct flock" and copy it to a "struct file_lock" as a POSIX
|
|
|
|
* style lock.
|
|
|
|
*/
|
|
|
|
static int flock_to_posix_lock(struct file *filp, struct file_lock *fl,
|
|
|
|
struct flock *l)
|
|
|
|
{
|
|
|
|
off_t start, end;
|
|
|
|
|
|
|
|
switch (l->l_whence) {
|
2006-10-01 00:27:35 -06:00
|
|
|
case SEEK_SET:
|
2005-04-16 16:20:36 -06:00
|
|
|
start = 0;
|
|
|
|
break;
|
2006-10-01 00:27:35 -06:00
|
|
|
case SEEK_CUR:
|
2005-04-16 16:20:36 -06:00
|
|
|
start = filp->f_pos;
|
|
|
|
break;
|
2006-10-01 00:27:35 -06:00
|
|
|
case SEEK_END:
|
2006-12-08 03:36:35 -07:00
|
|
|
start = i_size_read(filp->f_path.dentry->d_inode);
|
2005-04-16 16:20:36 -06:00
|
|
|
break;
|
|
|
|
default:
|
|
|
|
return -EINVAL;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* POSIX-1996 leaves the case l->l_len < 0 undefined;
|
|
|
|
POSIX-2001 defines it. */
|
|
|
|
start += l->l_start;
|
2005-10-18 15:20:21 -06:00
|
|
|
if (start < 0)
|
|
|
|
return -EINVAL;
|
|
|
|
fl->fl_end = OFFSET_MAX;
|
|
|
|
if (l->l_len > 0) {
|
|
|
|
end = start + l->l_len - 1;
|
|
|
|
fl->fl_end = end;
|
|
|
|
} else if (l->l_len < 0) {
|
2005-04-16 16:20:36 -06:00
|
|
|
end = start - 1;
|
2005-10-18 15:20:21 -06:00
|
|
|
fl->fl_end = end;
|
2005-04-16 16:20:36 -06:00
|
|
|
start += l->l_len;
|
2005-10-18 15:20:21 -06:00
|
|
|
if (start < 0)
|
|
|
|
return -EINVAL;
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
fl->fl_start = start; /* we record the absolute position */
|
2005-10-18 15:20:21 -06:00
|
|
|
if (fl->fl_end < fl->fl_start)
|
|
|
|
return -EOVERFLOW;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
fl->fl_owner = current->files;
|
|
|
|
fl->fl_pid = current->tgid;
|
|
|
|
fl->fl_file = filp;
|
|
|
|
fl->fl_flags = FL_POSIX;
|
|
|
|
fl->fl_ops = NULL;
|
|
|
|
fl->fl_lmops = NULL;
|
|
|
|
|
|
|
|
return assign_type(fl, l->l_type);
|
|
|
|
}
|
|
|
|
|
|
|
|
#if BITS_PER_LONG == 32
|
|
|
|
static int flock64_to_posix_lock(struct file *filp, struct file_lock *fl,
|
|
|
|
struct flock64 *l)
|
|
|
|
{
|
|
|
|
loff_t start;
|
|
|
|
|
|
|
|
switch (l->l_whence) {
|
2006-10-01 00:27:35 -06:00
|
|
|
case SEEK_SET:
|
2005-04-16 16:20:36 -06:00
|
|
|
start = 0;
|
|
|
|
break;
|
2006-10-01 00:27:35 -06:00
|
|
|
case SEEK_CUR:
|
2005-04-16 16:20:36 -06:00
|
|
|
start = filp->f_pos;
|
|
|
|
break;
|
2006-10-01 00:27:35 -06:00
|
|
|
case SEEK_END:
|
2006-12-08 03:36:35 -07:00
|
|
|
start = i_size_read(filp->f_path.dentry->d_inode);
|
2005-04-16 16:20:36 -06:00
|
|
|
break;
|
|
|
|
default:
|
|
|
|
return -EINVAL;
|
|
|
|
}
|
|
|
|
|
2005-10-18 15:20:21 -06:00
|
|
|
start += l->l_start;
|
|
|
|
if (start < 0)
|
2005-04-16 16:20:36 -06:00
|
|
|
return -EINVAL;
|
2005-10-18 15:20:21 -06:00
|
|
|
fl->fl_end = OFFSET_MAX;
|
|
|
|
if (l->l_len > 0) {
|
|
|
|
fl->fl_end = start + l->l_len - 1;
|
|
|
|
} else if (l->l_len < 0) {
|
|
|
|
fl->fl_end = start - 1;
|
|
|
|
start += l->l_len;
|
|
|
|
if (start < 0)
|
|
|
|
return -EINVAL;
|
|
|
|
}
|
2005-04-16 16:20:36 -06:00
|
|
|
fl->fl_start = start; /* we record the absolute position */
|
2005-10-18 15:20:21 -06:00
|
|
|
if (fl->fl_end < fl->fl_start)
|
|
|
|
return -EOVERFLOW;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
fl->fl_owner = current->files;
|
|
|
|
fl->fl_pid = current->tgid;
|
|
|
|
fl->fl_file = filp;
|
|
|
|
fl->fl_flags = FL_POSIX;
|
|
|
|
fl->fl_ops = NULL;
|
|
|
|
fl->fl_lmops = NULL;
|
|
|
|
|
2011-01-16 23:45:59 -07:00
|
|
|
return assign_type(fl, l->l_type);
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
|
|
|
/* default lease lock manager operations */
|
|
|
|
static void lease_break_callback(struct file_lock *fl)
|
|
|
|
{
|
|
|
|
kill_fasync(&fl->fl_fasync, SIGIO, POLL_MSG);
|
|
|
|
}
|
|
|
|
|
2009-09-21 18:01:12 -06:00
|
|
|
static const struct lock_manager_operations lease_manager_ops = {
|
2011-07-20 18:21:59 -06:00
|
|
|
.lm_break = lease_break_callback,
|
|
|
|
.lm_change = lease_modify,
|
2005-04-16 16:20:36 -06:00
|
|
|
};
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Initialize a lease, use the default lock manager operations
|
|
|
|
*/
|
2012-07-23 13:17:17 -06:00
|
|
|
static int lease_init(struct file *filp, long type, struct file_lock *fl)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
2006-05-07 21:02:42 -06:00
|
|
|
if (assign_type(fl, type) != 0)
|
|
|
|
return -EINVAL;
|
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
fl->fl_owner = current->files;
|
|
|
|
fl->fl_pid = current->tgid;
|
|
|
|
|
|
|
|
fl->fl_file = filp;
|
|
|
|
fl->fl_flags = FL_LEASE;
|
|
|
|
fl->fl_start = 0;
|
|
|
|
fl->fl_end = OFFSET_MAX;
|
|
|
|
fl->fl_ops = NULL;
|
|
|
|
fl->fl_lmops = &lease_manager_ops;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Allocate a file_lock initialised to this type of lease */
|
2012-07-23 13:17:17 -06:00
|
|
|
static struct file_lock *lease_alloc(struct file *filp, long type)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
|
|
|
struct file_lock *fl = locks_alloc_lock();
|
2006-05-07 21:02:42 -06:00
|
|
|
int error = -ENOMEM;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
if (fl == NULL)
|
2007-03-01 12:34:35 -07:00
|
|
|
return ERR_PTR(error);
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
error = lease_init(filp, type, fl);
|
2006-05-07 21:02:42 -06:00
|
|
|
if (error) {
|
|
|
|
locks_free_lock(fl);
|
2007-03-01 12:34:35 -07:00
|
|
|
return ERR_PTR(error);
|
2006-05-07 21:02:42 -06:00
|
|
|
}
|
2007-03-01 12:34:35 -07:00
|
|
|
return fl;
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
|
|
|
|
/* Check if two locks overlap each other.
|
|
|
|
*/
|
|
|
|
static inline int locks_overlap(struct file_lock *fl1, struct file_lock *fl2)
|
|
|
|
{
|
|
|
|
return ((fl1->fl_end >= fl2->fl_start) &&
|
|
|
|
(fl2->fl_end >= fl1->fl_start));
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Check whether two locks have the same owner.
|
|
|
|
*/
|
2006-01-08 02:05:22 -07:00
|
|
|
static int posix_same_owner(struct file_lock *fl1, struct file_lock *fl2)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
2011-07-20 18:21:59 -06:00
|
|
|
if (fl1->fl_lmops && fl1->fl_lmops->lm_compare_owner)
|
2005-04-16 16:20:36 -06:00
|
|
|
return fl2->fl_lmops == fl1->fl_lmops &&
|
2011-07-20 18:21:59 -06:00
|
|
|
fl1->fl_lmops->lm_compare_owner(fl1, fl2);
|
2005-04-16 16:20:36 -06:00
|
|
|
return fl1->fl_owner == fl2->fl_owner;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Remove waiter from blocker's block list.
|
|
|
|
* When blocker ends up pointing to itself then the list is empty.
|
|
|
|
*/
|
2006-01-08 02:05:22 -07:00
|
|
|
static void __locks_delete_block(struct file_lock *waiter)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
|
|
|
list_del_init(&waiter->fl_block);
|
|
|
|
list_del_init(&waiter->fl_link);
|
|
|
|
waiter->fl_next = NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
*/
|
2012-03-28 11:56:19 -06:00
|
|
|
void locks_delete_block(struct file_lock *waiter)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
2010-09-18 07:09:31 -06:00
|
|
|
lock_flocks();
|
2005-04-16 16:20:36 -06:00
|
|
|
__locks_delete_block(waiter);
|
2010-09-18 07:09:31 -06:00
|
|
|
unlock_flocks();
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
2012-03-28 11:56:19 -06:00
|
|
|
EXPORT_SYMBOL(locks_delete_block);
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
/* Insert waiter into blocker's block list.
|
|
|
|
* We use a circular list so that processes can be easily woken up in
|
|
|
|
* the order they blocked. The documentation doesn't require this but
|
|
|
|
* it seems like the reasonable thing to do.
|
|
|
|
*/
|
|
|
|
static void locks_insert_block(struct file_lock *blocker,
|
|
|
|
struct file_lock *waiter)
|
|
|
|
{
|
2006-03-26 02:37:24 -07:00
|
|
|
BUG_ON(!list_empty(&waiter->fl_block));
|
2005-04-16 16:20:36 -06:00
|
|
|
list_add_tail(&waiter->fl_block, &blocker->fl_block);
|
|
|
|
waiter->fl_next = blocker;
|
|
|
|
if (IS_POSIX(blocker))
|
|
|
|
list_add(&waiter->fl_link, &blocked_list);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Wake up processes blocked waiting for blocker.
|
|
|
|
* If told to wait then schedule the processes until the block list
|
|
|
|
* is empty, otherwise empty the block list ourselves.
|
|
|
|
*/
|
|
|
|
static void locks_wake_up_blocks(struct file_lock *blocker)
|
|
|
|
{
|
|
|
|
while (!list_empty(&blocker->fl_block)) {
|
2007-09-19 06:44:07 -06:00
|
|
|
struct file_lock *waiter;
|
|
|
|
|
|
|
|
waiter = list_first_entry(&blocker->fl_block,
|
2005-04-16 16:20:36 -06:00
|
|
|
struct file_lock, fl_block);
|
|
|
|
__locks_delete_block(waiter);
|
2011-07-20 18:21:59 -06:00
|
|
|
if (waiter->fl_lmops && waiter->fl_lmops->lm_notify)
|
|
|
|
waiter->fl_lmops->lm_notify(waiter);
|
2005-04-16 16:20:36 -06:00
|
|
|
else
|
|
|
|
wake_up(&waiter->fl_wait);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Insert file lock fl into an inode's lock list at the position indicated
|
|
|
|
* by pos. At the same time add the lock to the global file lock list.
|
|
|
|
*/
|
|
|
|
static void locks_insert_lock(struct file_lock **pos, struct file_lock *fl)
|
|
|
|
{
|
|
|
|
list_add(&fl->fl_link, &file_lock_list);
|
|
|
|
|
2008-01-16 17:07:08 -07:00
|
|
|
fl->fl_nspid = get_pid(task_tgid(current));
|
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
/* insert into file's list */
|
|
|
|
fl->fl_next = *pos;
|
|
|
|
*pos = fl;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Delete a lock and then free it.
|
|
|
|
* Wake up processes that are blocked waiting for this lock,
|
|
|
|
* notify the FS that the lock has been cleared and
|
|
|
|
* finally free the lock.
|
|
|
|
*/
|
|
|
|
static void locks_delete_lock(struct file_lock **thisfl_p)
|
|
|
|
{
|
|
|
|
struct file_lock *fl = *thisfl_p;
|
|
|
|
|
|
|
|
*thisfl_p = fl->fl_next;
|
|
|
|
fl->fl_next = NULL;
|
|
|
|
list_del_init(&fl->fl_link);
|
|
|
|
|
2008-01-16 17:07:08 -07:00
|
|
|
if (fl->fl_nspid) {
|
|
|
|
put_pid(fl->fl_nspid);
|
|
|
|
fl->fl_nspid = NULL;
|
|
|
|
}
|
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
locks_wake_up_blocks(fl);
|
|
|
|
locks_free_lock(fl);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Determine if lock sys_fl blocks lock caller_fl. Common functionality
|
|
|
|
* checks for shared/exclusive status of overlapping locks.
|
|
|
|
*/
|
|
|
|
static int locks_conflict(struct file_lock *caller_fl, struct file_lock *sys_fl)
|
|
|
|
{
|
|
|
|
if (sys_fl->fl_type == F_WRLCK)
|
|
|
|
return 1;
|
|
|
|
if (caller_fl->fl_type == F_WRLCK)
|
|
|
|
return 1;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Determine if lock sys_fl blocks lock caller_fl. POSIX specific
|
|
|
|
* checking before calling the locks_conflict().
|
|
|
|
*/
|
|
|
|
static int posix_locks_conflict(struct file_lock *caller_fl, struct file_lock *sys_fl)
|
|
|
|
{
|
|
|
|
/* POSIX locks owned by the same process do not conflict with
|
|
|
|
* each other.
|
|
|
|
*/
|
|
|
|
if (!IS_POSIX(sys_fl) || posix_same_owner(caller_fl, sys_fl))
|
|
|
|
return (0);
|
|
|
|
|
|
|
|
/* Check whether they overlap */
|
|
|
|
if (!locks_overlap(caller_fl, sys_fl))
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
return (locks_conflict(caller_fl, sys_fl));
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Determine if lock sys_fl blocks lock caller_fl. FLOCK specific
|
|
|
|
* checking before calling the locks_conflict().
|
|
|
|
*/
|
|
|
|
static int flock_locks_conflict(struct file_lock *caller_fl, struct file_lock *sys_fl)
|
|
|
|
{
|
|
|
|
/* FLOCK locks referring to the same filp do not conflict with
|
|
|
|
* each other.
|
|
|
|
*/
|
|
|
|
if (!IS_FLOCK(sys_fl) || (caller_fl->fl_file == sys_fl->fl_file))
|
|
|
|
return (0);
|
|
|
|
if ((caller_fl->fl_type & LOCK_MAND) || (sys_fl->fl_type & LOCK_MAND))
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
return (locks_conflict(caller_fl, sys_fl));
|
|
|
|
}
|
|
|
|
|
2007-05-11 14:09:32 -06:00
|
|
|
void
|
2007-02-20 22:55:18 -07:00
|
|
|
posix_test_lock(struct file *filp, struct file_lock *fl)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
|
|
|
struct file_lock *cfl;
|
|
|
|
|
2010-09-18 07:09:31 -06:00
|
|
|
lock_flocks();
|
2006-12-08 03:36:35 -07:00
|
|
|
for (cfl = filp->f_path.dentry->d_inode->i_flock; cfl; cfl = cfl->fl_next) {
|
2005-04-16 16:20:36 -06:00
|
|
|
if (!IS_POSIX(cfl))
|
|
|
|
continue;
|
2007-05-10 17:02:07 -06:00
|
|
|
if (posix_locks_conflict(fl, cfl))
|
2005-04-16 16:20:36 -06:00
|
|
|
break;
|
|
|
|
}
|
2008-01-16 17:07:08 -07:00
|
|
|
if (cfl) {
|
2007-02-20 22:55:18 -07:00
|
|
|
__locks_copy_lock(fl, cfl);
|
2008-01-16 17:07:08 -07:00
|
|
|
if (cfl->fl_nspid)
|
2008-02-08 05:19:20 -07:00
|
|
|
fl->fl_pid = pid_vnr(cfl->fl_nspid);
|
2008-01-16 17:07:08 -07:00
|
|
|
} else
|
2007-05-10 16:38:43 -06:00
|
|
|
fl->fl_type = F_UNLCK;
|
2010-09-18 07:09:31 -06:00
|
|
|
unlock_flocks();
|
2007-05-11 14:09:32 -06:00
|
|
|
return;
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(posix_test_lock);
|
|
|
|
|
2007-10-26 16:05:40 -06:00
|
|
|
/*
|
|
|
|
* Deadlock detection:
|
|
|
|
*
|
|
|
|
* We attempt to detect deadlocks that are due purely to posix file
|
|
|
|
* locks.
|
2005-04-16 16:20:36 -06:00
|
|
|
*
|
2007-10-26 16:05:40 -06:00
|
|
|
* We assume that a task can be waiting for at most one lock at a time.
|
|
|
|
* So for any acquired lock, the process holding that lock may be
|
|
|
|
* waiting on at most one other lock. That lock in turns may be held by
|
|
|
|
* someone waiting for at most one other lock. Given a requested lock
|
|
|
|
* caller_fl which is about to wait for a conflicting lock block_fl, we
|
|
|
|
* follow this chain of waiters to ensure we are not about to create a
|
|
|
|
* cycle.
|
2005-04-16 16:20:36 -06:00
|
|
|
*
|
2007-10-26 16:05:40 -06:00
|
|
|
* Since we do this before we ever put a process to sleep on a lock, we
|
|
|
|
* are ensured that there is never a cycle; that is what guarantees that
|
|
|
|
* the while() loop in posix_locks_deadlock() eventually completes.
|
2007-10-30 09:20:02 -06:00
|
|
|
*
|
2007-10-26 16:05:40 -06:00
|
|
|
* Note: the above assumption may not be true when handling lock
|
|
|
|
* requests from a broken NFS client. It may also fail in the presence
|
|
|
|
* of tasks (such as posix threads) sharing the same open file table.
|
|
|
|
*
|
|
|
|
* To handle those cases, we just bail out after a few iterations.
|
2005-04-16 16:20:36 -06:00
|
|
|
*/
|
2007-10-30 09:20:02 -06:00
|
|
|
|
|
|
|
#define MAX_DEADLK_ITERATIONS 10
|
|
|
|
|
2007-10-26 16:05:40 -06:00
|
|
|
/* Find a lock that the owner of the given block_fl is blocking on. */
|
|
|
|
static struct file_lock *what_owner_is_waiting_for(struct file_lock *block_fl)
|
|
|
|
{
|
|
|
|
struct file_lock *fl;
|
|
|
|
|
|
|
|
list_for_each_entry(fl, &blocked_list, fl_link) {
|
|
|
|
if (posix_same_owner(fl, block_fl))
|
|
|
|
return fl->fl_next;
|
|
|
|
}
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
2006-06-23 03:05:13 -06:00
|
|
|
static int posix_locks_deadlock(struct file_lock *caller_fl,
|
2005-04-16 16:20:36 -06:00
|
|
|
struct file_lock *block_fl)
|
|
|
|
{
|
2007-10-30 09:20:02 -06:00
|
|
|
int i = 0;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2007-10-26 16:05:40 -06:00
|
|
|
while ((block_fl = what_owner_is_waiting_for(block_fl))) {
|
|
|
|
if (i++ > MAX_DEADLK_ITERATIONS)
|
|
|
|
return 0;
|
|
|
|
if (posix_same_owner(caller_fl, block_fl))
|
|
|
|
return 1;
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Try to create a FLOCK lock on filp. We always insert new FLOCK locks
|
2007-09-12 13:45:07 -06:00
|
|
|
* after any leases, but before any posix locks.
|
2006-06-29 14:38:32 -06:00
|
|
|
*
|
|
|
|
* Note that if called with an FL_EXISTS argument, the caller may determine
|
|
|
|
* whether or not a lock was successfully freed by testing the return
|
|
|
|
* value for -ENOENT.
|
2005-04-16 16:20:36 -06:00
|
|
|
*/
|
2006-03-31 03:30:55 -07:00
|
|
|
static int flock_lock_file(struct file *filp, struct file_lock *request)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
2006-03-31 03:30:55 -07:00
|
|
|
struct file_lock *new_fl = NULL;
|
2005-04-16 16:20:36 -06:00
|
|
|
struct file_lock **before;
|
2006-12-08 03:36:35 -07:00
|
|
|
struct inode * inode = filp->f_path.dentry->d_inode;
|
2005-04-16 16:20:36 -06:00
|
|
|
int error = 0;
|
|
|
|
int found = 0;
|
|
|
|
|
2010-09-18 07:09:31 -06:00
|
|
|
if (!(request->fl_flags & FL_ACCESS) && (request->fl_type != F_UNLCK)) {
|
2007-09-11 06:38:13 -06:00
|
|
|
new_fl = locks_alloc_lock();
|
2010-09-18 07:09:31 -06:00
|
|
|
if (!new_fl)
|
|
|
|
return -ENOMEM;
|
2007-09-11 06:38:13 -06:00
|
|
|
}
|
|
|
|
|
2010-09-18 07:09:31 -06:00
|
|
|
lock_flocks();
|
|
|
|
if (request->fl_flags & FL_ACCESS)
|
|
|
|
goto find_conflict;
|
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
for_each_lock(inode, before) {
|
|
|
|
struct file_lock *fl = *before;
|
|
|
|
if (IS_POSIX(fl))
|
|
|
|
break;
|
|
|
|
if (IS_LEASE(fl))
|
|
|
|
continue;
|
|
|
|
if (filp != fl->fl_file)
|
|
|
|
continue;
|
2006-03-31 03:30:55 -07:00
|
|
|
if (request->fl_type == fl->fl_type)
|
2005-04-16 16:20:36 -06:00
|
|
|
goto out;
|
|
|
|
found = 1;
|
|
|
|
locks_delete_lock(before);
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
2006-06-29 14:38:32 -06:00
|
|
|
if (request->fl_type == F_UNLCK) {
|
|
|
|
if ((request->fl_flags & FL_EXISTS) && !found)
|
|
|
|
error = -ENOENT;
|
2006-03-31 03:30:55 -07:00
|
|
|
goto out;
|
2006-06-29 14:38:32 -06:00
|
|
|
}
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
/*
|
|
|
|
* If a higher-priority process was blocked on the old file lock,
|
|
|
|
* give it the opportunity to lock the file.
|
|
|
|
*/
|
2010-09-18 07:09:31 -06:00
|
|
|
if (found) {
|
|
|
|
unlock_flocks();
|
2009-07-16 07:44:29 -06:00
|
|
|
cond_resched();
|
2010-09-18 07:09:31 -06:00
|
|
|
lock_flocks();
|
|
|
|
}
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2006-06-29 14:38:37 -06:00
|
|
|
find_conflict:
|
2005-04-16 16:20:36 -06:00
|
|
|
for_each_lock(inode, before) {
|
|
|
|
struct file_lock *fl = *before;
|
|
|
|
if (IS_POSIX(fl))
|
|
|
|
break;
|
|
|
|
if (IS_LEASE(fl))
|
|
|
|
continue;
|
2006-03-31 03:30:55 -07:00
|
|
|
if (!flock_locks_conflict(request, fl))
|
2005-04-16 16:20:36 -06:00
|
|
|
continue;
|
|
|
|
error = -EAGAIN;
|
2008-07-25 02:48:57 -06:00
|
|
|
if (!(request->fl_flags & FL_SLEEP))
|
|
|
|
goto out;
|
|
|
|
error = FILE_LOCK_DEFERRED;
|
|
|
|
locks_insert_block(fl, request);
|
2005-04-16 16:20:36 -06:00
|
|
|
goto out;
|
|
|
|
}
|
2006-06-29 14:38:37 -06:00
|
|
|
if (request->fl_flags & FL_ACCESS)
|
|
|
|
goto out;
|
2006-03-31 03:30:55 -07:00
|
|
|
locks_copy_lock(new_fl, request);
|
Leases can be hidden by flocks
The inode->i_flock list contains the leases, flocks and posix
locks in the specified order. However, the flocks are added in
the head of this list thus hiding the leases from F_GETLEASE
command, from time_out_leases() and other code that expects
the leases to come first.
The following example will demonstrate this:
#define _GNU_SOURCE
#include <unistd.h>
#include <fcntl.h>
#include <stdio.h>
#include <sys/file.h>
static void show_lease(int fd)
{
int res;
res = fcntl(fd, F_GETLEASE);
switch (res) {
case F_RDLCK:
printf("Read lease\n");
break;
case F_WRLCK:
printf("Write lease\n");
break;
case F_UNLCK:
printf("No leases\n");
break;
default:
printf("Some shit\n");
break;
}
}
int main(int argc, char **argv)
{
int fd, res;
fd = open(argv[1], O_RDONLY);
if (fd == -1) {
perror("Can't open file");
return 1;
}
res = fcntl(fd, F_SETLEASE, F_WRLCK);
if (res == -1) {
perror("Can't set lease");
return 1;
}
show_lease(fd);
if (flock(fd, LOCK_SH) == -1) {
perror("Can't flock shared");
return 1;
}
show_lease(fd);
return 0;
}
The first call to show_lease() will show the write lease set, but
the second will show no leases.
Fix the flock adding so that the leases always stay in the head
of this list.
Found during making the flocks pid-namespaces aware.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Acked-by: "J. Bruce Fields" <bfields@fieldses.org>
Cc: Trond Myklebust <trond.myklebust@fys.uio.no>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: <stable@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-09-11 16:24:01 -06:00
|
|
|
locks_insert_lock(before, new_fl);
|
2006-03-31 03:30:55 -07:00
|
|
|
new_fl = NULL;
|
2006-06-14 07:59:35 -06:00
|
|
|
error = 0;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
out:
|
2010-09-18 07:09:31 -06:00
|
|
|
unlock_flocks();
|
2006-03-31 03:30:55 -07:00
|
|
|
if (new_fl)
|
|
|
|
locks_free_lock(new_fl);
|
2005-04-16 16:20:36 -06:00
|
|
|
return error;
|
|
|
|
}
|
|
|
|
|
2007-01-18 14:15:35 -07:00
|
|
|
static int __posix_lock_file(struct inode *inode, struct file_lock *request, struct file_lock *conflock)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
|
|
|
struct file_lock *fl;
|
2006-06-23 03:05:10 -06:00
|
|
|
struct file_lock *new_fl = NULL;
|
|
|
|
struct file_lock *new_fl2 = NULL;
|
2005-04-16 16:20:36 -06:00
|
|
|
struct file_lock *left = NULL;
|
|
|
|
struct file_lock *right = NULL;
|
|
|
|
struct file_lock **before;
|
|
|
|
int error, added = 0;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* We may need two file_lock structures for this operation,
|
|
|
|
* so we get them in advance to avoid races.
|
2006-06-23 03:05:10 -06:00
|
|
|
*
|
|
|
|
* In some cases we can be sure, that no new locks will be needed
|
2005-04-16 16:20:36 -06:00
|
|
|
*/
|
2006-06-23 03:05:10 -06:00
|
|
|
if (!(request->fl_flags & FL_ACCESS) &&
|
|
|
|
(request->fl_type != F_UNLCK ||
|
|
|
|
request->fl_start != 0 || request->fl_end != OFFSET_MAX)) {
|
|
|
|
new_fl = locks_alloc_lock();
|
|
|
|
new_fl2 = locks_alloc_lock();
|
|
|
|
}
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2010-09-18 07:09:31 -06:00
|
|
|
lock_flocks();
|
2005-04-16 16:20:36 -06:00
|
|
|
if (request->fl_type != F_UNLCK) {
|
|
|
|
for_each_lock(inode, before) {
|
2006-11-14 14:54:36 -07:00
|
|
|
fl = *before;
|
2005-04-16 16:20:36 -06:00
|
|
|
if (!IS_POSIX(fl))
|
|
|
|
continue;
|
|
|
|
if (!posix_locks_conflict(request, fl))
|
|
|
|
continue;
|
2006-03-26 02:37:26 -07:00
|
|
|
if (conflock)
|
2008-04-24 08:08:22 -06:00
|
|
|
__locks_copy_lock(conflock, fl);
|
2005-04-16 16:20:36 -06:00
|
|
|
error = -EAGAIN;
|
|
|
|
if (!(request->fl_flags & FL_SLEEP))
|
|
|
|
goto out;
|
|
|
|
error = -EDEADLK;
|
|
|
|
if (posix_locks_deadlock(request, fl))
|
|
|
|
goto out;
|
2008-07-25 02:48:57 -06:00
|
|
|
error = FILE_LOCK_DEFERRED;
|
2005-04-16 16:20:36 -06:00
|
|
|
locks_insert_block(fl, request);
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* If we're just looking for a conflict, we're done. */
|
|
|
|
error = 0;
|
|
|
|
if (request->fl_flags & FL_ACCESS)
|
|
|
|
goto out;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Find the first old lock with the same owner as the new lock.
|
|
|
|
*/
|
|
|
|
|
|
|
|
before = &inode->i_flock;
|
|
|
|
|
|
|
|
/* First skip locks owned by other processes. */
|
|
|
|
while ((fl = *before) && (!IS_POSIX(fl) ||
|
|
|
|
!posix_same_owner(request, fl))) {
|
|
|
|
before = &fl->fl_next;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Process locks with this owner. */
|
|
|
|
while ((fl = *before) && posix_same_owner(request, fl)) {
|
|
|
|
/* Detect adjacent or overlapping regions (if same lock type)
|
|
|
|
*/
|
|
|
|
if (request->fl_type == fl->fl_type) {
|
2005-08-25 17:25:35 -06:00
|
|
|
/* In all comparisons of start vs end, use
|
|
|
|
* "start - 1" rather than "end + 1". If end
|
|
|
|
* is OFFSET_MAX, end + 1 will become negative.
|
|
|
|
*/
|
2005-04-16 16:20:36 -06:00
|
|
|
if (fl->fl_end < request->fl_start - 1)
|
|
|
|
goto next_lock;
|
|
|
|
/* If the next lock in the list has entirely bigger
|
|
|
|
* addresses than the new one, insert the lock here.
|
|
|
|
*/
|
2005-08-25 17:25:35 -06:00
|
|
|
if (fl->fl_start - 1 > request->fl_end)
|
2005-04-16 16:20:36 -06:00
|
|
|
break;
|
|
|
|
|
|
|
|
/* If we come here, the new and old lock are of the
|
|
|
|
* same type and adjacent or overlapping. Make one
|
|
|
|
* lock yielding from the lower start address of both
|
|
|
|
* locks to the higher end address.
|
|
|
|
*/
|
|
|
|
if (fl->fl_start > request->fl_start)
|
|
|
|
fl->fl_start = request->fl_start;
|
|
|
|
else
|
|
|
|
request->fl_start = fl->fl_start;
|
|
|
|
if (fl->fl_end < request->fl_end)
|
|
|
|
fl->fl_end = request->fl_end;
|
|
|
|
else
|
|
|
|
request->fl_end = fl->fl_end;
|
|
|
|
if (added) {
|
|
|
|
locks_delete_lock(before);
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
request = fl;
|
|
|
|
added = 1;
|
|
|
|
}
|
|
|
|
else {
|
|
|
|
/* Processing for different lock types is a bit
|
|
|
|
* more complex.
|
|
|
|
*/
|
|
|
|
if (fl->fl_end < request->fl_start)
|
|
|
|
goto next_lock;
|
|
|
|
if (fl->fl_start > request->fl_end)
|
|
|
|
break;
|
|
|
|
if (request->fl_type == F_UNLCK)
|
|
|
|
added = 1;
|
|
|
|
if (fl->fl_start < request->fl_start)
|
|
|
|
left = fl;
|
|
|
|
/* If the next lock in the list has a higher end
|
|
|
|
* address than the new one, insert the new one here.
|
|
|
|
*/
|
|
|
|
if (fl->fl_end > request->fl_end) {
|
|
|
|
right = fl;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
if (fl->fl_start >= request->fl_start) {
|
|
|
|
/* The new lock completely replaces an old
|
|
|
|
* one (This may happen several times).
|
|
|
|
*/
|
|
|
|
if (added) {
|
|
|
|
locks_delete_lock(before);
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
/* Replace the old lock with the new one.
|
|
|
|
* Wake up anybody waiting for the old one,
|
|
|
|
* as the change in lock type might satisfy
|
|
|
|
* their needs.
|
|
|
|
*/
|
|
|
|
locks_wake_up_blocks(fl);
|
|
|
|
fl->fl_start = request->fl_start;
|
|
|
|
fl->fl_end = request->fl_end;
|
|
|
|
fl->fl_type = request->fl_type;
|
2006-03-20 11:44:05 -07:00
|
|
|
locks_release_private(fl);
|
|
|
|
locks_copy_private(fl, request);
|
2005-04-16 16:20:36 -06:00
|
|
|
request = fl;
|
|
|
|
added = 1;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
/* Go on to next lock.
|
|
|
|
*/
|
|
|
|
next_lock:
|
|
|
|
before = &fl->fl_next;
|
|
|
|
}
|
|
|
|
|
2006-06-23 03:05:09 -06:00
|
|
|
/*
|
|
|
|
* The above code only modifies existing locks in case of
|
|
|
|
* merging or replacing. If new lock(s) need to be inserted
|
|
|
|
* all modifications are done bellow this, so it's safe yet to
|
|
|
|
* bail out.
|
|
|
|
*/
|
|
|
|
error = -ENOLCK; /* "no luck" */
|
|
|
|
if (right && left == right && !new_fl2)
|
|
|
|
goto out;
|
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
error = 0;
|
|
|
|
if (!added) {
|
2006-06-29 14:38:32 -06:00
|
|
|
if (request->fl_type == F_UNLCK) {
|
|
|
|
if (request->fl_flags & FL_EXISTS)
|
|
|
|
error = -ENOENT;
|
2005-04-16 16:20:36 -06:00
|
|
|
goto out;
|
2006-06-29 14:38:32 -06:00
|
|
|
}
|
2006-06-23 03:05:09 -06:00
|
|
|
|
|
|
|
if (!new_fl) {
|
|
|
|
error = -ENOLCK;
|
|
|
|
goto out;
|
|
|
|
}
|
2005-04-16 16:20:36 -06:00
|
|
|
locks_copy_lock(new_fl, request);
|
|
|
|
locks_insert_lock(before, new_fl);
|
|
|
|
new_fl = NULL;
|
|
|
|
}
|
|
|
|
if (right) {
|
|
|
|
if (left == right) {
|
|
|
|
/* The new lock breaks the old one in two pieces,
|
|
|
|
* so we have to use the second new lock.
|
|
|
|
*/
|
|
|
|
left = new_fl2;
|
|
|
|
new_fl2 = NULL;
|
|
|
|
locks_copy_lock(left, right);
|
|
|
|
locks_insert_lock(before, left);
|
|
|
|
}
|
|
|
|
right->fl_start = request->fl_end + 1;
|
|
|
|
locks_wake_up_blocks(right);
|
|
|
|
}
|
|
|
|
if (left) {
|
|
|
|
left->fl_end = request->fl_start - 1;
|
|
|
|
locks_wake_up_blocks(left);
|
|
|
|
}
|
|
|
|
out:
|
2010-09-18 07:09:31 -06:00
|
|
|
unlock_flocks();
|
2005-04-16 16:20:36 -06:00
|
|
|
/*
|
|
|
|
* Free any unused locks.
|
|
|
|
*/
|
|
|
|
if (new_fl)
|
|
|
|
locks_free_lock(new_fl);
|
|
|
|
if (new_fl2)
|
|
|
|
locks_free_lock(new_fl2);
|
|
|
|
return error;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* posix_lock_file - Apply a POSIX-style lock to a file
|
|
|
|
* @filp: The file to apply the lock to
|
|
|
|
* @fl: The lock to be applied
|
2007-01-18 14:15:35 -07:00
|
|
|
* @conflock: Place to return a copy of the conflicting lock, if found.
|
2005-04-16 16:20:36 -06:00
|
|
|
*
|
|
|
|
* Add a POSIX style lock to a file.
|
|
|
|
* We merge adjacent & overlapping locks whenever possible.
|
|
|
|
* POSIX locks are sorted by owner task, then by starting address
|
2006-06-29 14:38:32 -06:00
|
|
|
*
|
|
|
|
* Note that if called with an FL_EXISTS argument, the caller may determine
|
|
|
|
* whether or not a lock was successfully freed by testing the return
|
|
|
|
* value for -ENOENT.
|
2005-04-16 16:20:36 -06:00
|
|
|
*/
|
2007-01-18 14:15:35 -07:00
|
|
|
int posix_lock_file(struct file *filp, struct file_lock *fl,
|
2006-03-26 02:37:26 -07:00
|
|
|
struct file_lock *conflock)
|
|
|
|
{
|
2007-01-18 14:15:35 -07:00
|
|
|
return __posix_lock_file(filp->f_path.dentry->d_inode, fl, conflock);
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
2007-01-18 14:15:35 -07:00
|
|
|
EXPORT_SYMBOL(posix_lock_file);
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
/**
|
|
|
|
* posix_lock_file_wait - Apply a POSIX-style lock to a file
|
|
|
|
* @filp: The file to apply the lock to
|
|
|
|
* @fl: The lock to be applied
|
|
|
|
*
|
|
|
|
* Add a POSIX style lock to a file.
|
|
|
|
* We merge adjacent & overlapping locks whenever possible.
|
|
|
|
* POSIX locks are sorted by owner task, then by starting address
|
|
|
|
*/
|
|
|
|
int posix_lock_file_wait(struct file *filp, struct file_lock *fl)
|
|
|
|
{
|
|
|
|
int error;
|
|
|
|
might_sleep ();
|
|
|
|
for (;;) {
|
2007-01-18 14:15:35 -07:00
|
|
|
error = posix_lock_file(filp, fl, NULL);
|
2008-07-25 02:48:57 -06:00
|
|
|
if (error != FILE_LOCK_DEFERRED)
|
2005-04-16 16:20:36 -06:00
|
|
|
break;
|
|
|
|
error = wait_event_interruptible(fl->fl_wait, !fl->fl_next);
|
|
|
|
if (!error)
|
|
|
|
continue;
|
|
|
|
|
|
|
|
locks_delete_block(fl);
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
return error;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(posix_lock_file_wait);
|
|
|
|
|
|
|
|
/**
|
|
|
|
* locks_mandatory_locked - Check for an active lock
|
|
|
|
* @inode: the file to check
|
|
|
|
*
|
|
|
|
* Searches the inode's list of locks to find any POSIX locks which conflict.
|
|
|
|
* This function is called from locks_verify_locked() only.
|
|
|
|
*/
|
|
|
|
int locks_mandatory_locked(struct inode *inode)
|
|
|
|
{
|
|
|
|
fl_owner_t owner = current->files;
|
|
|
|
struct file_lock *fl;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Search the lock list for this inode for any POSIX locks.
|
|
|
|
*/
|
2010-09-18 07:09:31 -06:00
|
|
|
lock_flocks();
|
2005-04-16 16:20:36 -06:00
|
|
|
for (fl = inode->i_flock; fl != NULL; fl = fl->fl_next) {
|
|
|
|
if (!IS_POSIX(fl))
|
|
|
|
continue;
|
|
|
|
if (fl->fl_owner != owner)
|
|
|
|
break;
|
|
|
|
}
|
2010-09-18 07:09:31 -06:00
|
|
|
unlock_flocks();
|
2005-04-16 16:20:36 -06:00
|
|
|
return fl ? -EAGAIN : 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* locks_mandatory_area - Check for a conflicting lock
|
|
|
|
* @read_write: %FLOCK_VERIFY_WRITE for exclusive access, %FLOCK_VERIFY_READ
|
|
|
|
* for shared
|
|
|
|
* @inode: the file to check
|
|
|
|
* @filp: how the file was opened (if it was)
|
|
|
|
* @offset: start of area to check
|
|
|
|
* @count: length of area to check
|
|
|
|
*
|
|
|
|
* Searches the inode's list of locks to find any POSIX locks which conflict.
|
|
|
|
* This function is called from rw_verify_area() and
|
|
|
|
* locks_verify_truncate().
|
|
|
|
*/
|
|
|
|
int locks_mandatory_area(int read_write, struct inode *inode,
|
|
|
|
struct file *filp, loff_t offset,
|
|
|
|
size_t count)
|
|
|
|
{
|
|
|
|
struct file_lock fl;
|
|
|
|
int error;
|
|
|
|
|
|
|
|
locks_init_lock(&fl);
|
|
|
|
fl.fl_owner = current->files;
|
|
|
|
fl.fl_pid = current->tgid;
|
|
|
|
fl.fl_file = filp;
|
|
|
|
fl.fl_flags = FL_POSIX | FL_ACCESS;
|
|
|
|
if (filp && !(filp->f_flags & O_NONBLOCK))
|
|
|
|
fl.fl_flags |= FL_SLEEP;
|
|
|
|
fl.fl_type = (read_write == FLOCK_VERIFY_WRITE) ? F_WRLCK : F_RDLCK;
|
|
|
|
fl.fl_start = offset;
|
|
|
|
fl.fl_end = offset + count - 1;
|
|
|
|
|
|
|
|
for (;;) {
|
2007-01-18 14:15:35 -07:00
|
|
|
error = __posix_lock_file(inode, &fl, NULL);
|
2008-07-25 02:48:57 -06:00
|
|
|
if (error != FILE_LOCK_DEFERRED)
|
2005-04-16 16:20:36 -06:00
|
|
|
break;
|
|
|
|
error = wait_event_interruptible(fl.fl_wait, !fl.fl_next);
|
|
|
|
if (!error) {
|
|
|
|
/*
|
|
|
|
* If we've been sleeping someone might have
|
|
|
|
* changed the permissions behind our back.
|
|
|
|
*/
|
2007-10-01 15:41:11 -06:00
|
|
|
if (__mandatory_lock(inode))
|
2005-04-16 16:20:36 -06:00
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
|
|
|
locks_delete_block(&fl);
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
return error;
|
|
|
|
}
|
|
|
|
|
|
|
|
EXPORT_SYMBOL(locks_mandatory_area);
|
|
|
|
|
2011-07-26 16:25:49 -06:00
|
|
|
static void lease_clear_pending(struct file_lock *fl, int arg)
|
|
|
|
{
|
|
|
|
switch (arg) {
|
|
|
|
case F_UNLCK:
|
|
|
|
fl->fl_flags &= ~FL_UNLOCK_PENDING;
|
|
|
|
/* fall through: */
|
|
|
|
case F_RDLCK:
|
|
|
|
fl->fl_flags &= ~FL_DOWNGRADE_PENDING;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
/* We already had a lease on this file; just change its type */
|
|
|
|
int lease_modify(struct file_lock **before, int arg)
|
|
|
|
{
|
|
|
|
struct file_lock *fl = *before;
|
|
|
|
int error = assign_type(fl, arg);
|
|
|
|
|
|
|
|
if (error)
|
|
|
|
return error;
|
2011-07-26 16:25:49 -06:00
|
|
|
lease_clear_pending(fl, arg);
|
2005-04-16 16:20:36 -06:00
|
|
|
locks_wake_up_blocks(fl);
|
2012-07-26 22:42:52 -06:00
|
|
|
if (arg == F_UNLCK) {
|
|
|
|
struct file *filp = fl->fl_file;
|
|
|
|
|
|
|
|
f_delown(filp);
|
|
|
|
filp->f_owner.signum = 0;
|
2012-07-27 14:18:00 -06:00
|
|
|
fasync_helper(0, fl->fl_file, 0, &fl->fl_fasync);
|
|
|
|
if (fl->fl_fasync != NULL) {
|
|
|
|
printk(KERN_ERR "locks_delete_lock: fasync == %p\n", fl->fl_fasync);
|
|
|
|
fl->fl_fasync = NULL;
|
|
|
|
}
|
2005-04-16 16:20:36 -06:00
|
|
|
locks_delete_lock(before);
|
2012-07-26 22:42:52 -06:00
|
|
|
}
|
2005-04-16 16:20:36 -06:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
EXPORT_SYMBOL(lease_modify);
|
|
|
|
|
2011-07-26 16:25:49 -06:00
|
|
|
static bool past_time(unsigned long then)
|
|
|
|
{
|
|
|
|
if (!then)
|
|
|
|
/* 0 is a special value meaning "this never expires": */
|
|
|
|
return false;
|
|
|
|
return time_after(jiffies, then);
|
|
|
|
}
|
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
static void time_out_leases(struct inode *inode)
|
|
|
|
{
|
|
|
|
struct file_lock **before;
|
|
|
|
struct file_lock *fl;
|
|
|
|
|
|
|
|
before = &inode->i_flock;
|
2011-07-26 18:10:51 -06:00
|
|
|
while ((fl = *before) && IS_LEASE(fl) && lease_breaking(fl)) {
|
2011-07-26 16:25:49 -06:00
|
|
|
if (past_time(fl->fl_downgrade_time))
|
|
|
|
lease_modify(before, F_RDLCK);
|
|
|
|
if (past_time(fl->fl_break_time))
|
|
|
|
lease_modify(before, F_UNLCK);
|
2005-04-16 16:20:36 -06:00
|
|
|
if (fl == *before) /* lease_modify may have freed fl */
|
|
|
|
before = &fl->fl_next;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* __break_lease - revoke all outstanding leases on file
|
|
|
|
* @inode: the inode of the file to return
|
|
|
|
* @mode: the open mode (read or write)
|
|
|
|
*
|
2007-05-09 14:10:27 -06:00
|
|
|
* break_lease (inlined for speed) has checked there already is at least
|
|
|
|
* some kind of lock (maybe a lease) on this file. Leases are broken on
|
|
|
|
* a call to open() or truncate(). This function can sleep unless you
|
2005-04-16 16:20:36 -06:00
|
|
|
* specified %O_NONBLOCK to your open().
|
|
|
|
*/
|
|
|
|
int __break_lease(struct inode *inode, unsigned int mode)
|
|
|
|
{
|
2011-07-26 16:25:49 -06:00
|
|
|
int error = 0;
|
2005-04-16 16:20:36 -06:00
|
|
|
struct file_lock *new_fl, *flock;
|
|
|
|
struct file_lock *fl;
|
|
|
|
unsigned long break_time;
|
|
|
|
int i_have_this_lease = 0;
|
2009-12-24 04:47:55 -07:00
|
|
|
int want_write = (mode & O_ACCMODE) != O_RDONLY;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2009-12-24 04:47:55 -07:00
|
|
|
new_fl = lease_alloc(NULL, want_write ? F_WRLCK : F_RDLCK);
|
2011-12-26 11:25:26 -07:00
|
|
|
if (IS_ERR(new_fl))
|
|
|
|
return PTR_ERR(new_fl);
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2010-09-18 07:09:31 -06:00
|
|
|
lock_flocks();
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
time_out_leases(inode);
|
|
|
|
|
|
|
|
flock = inode->i_flock;
|
|
|
|
if ((flock == NULL) || !IS_LEASE(flock))
|
|
|
|
goto out;
|
|
|
|
|
2011-07-26 16:25:49 -06:00
|
|
|
if (!locks_conflict(flock, new_fl))
|
|
|
|
goto out;
|
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
for (fl = flock; fl && IS_LEASE(fl); fl = fl->fl_next)
|
|
|
|
if (fl->fl_owner == current->files)
|
|
|
|
i_have_this_lease = 1;
|
|
|
|
|
|
|
|
break_time = 0;
|
|
|
|
if (lease_break_time > 0) {
|
|
|
|
break_time = jiffies + lease_break_time * HZ;
|
|
|
|
if (break_time == 0)
|
|
|
|
break_time++; /* so that 0 means no break time */
|
|
|
|
}
|
|
|
|
|
|
|
|
for (fl = flock; fl && IS_LEASE(fl); fl = fl->fl_next) {
|
2011-07-26 16:25:49 -06:00
|
|
|
if (want_write) {
|
|
|
|
if (fl->fl_flags & FL_UNLOCK_PENDING)
|
|
|
|
continue;
|
|
|
|
fl->fl_flags |= FL_UNLOCK_PENDING;
|
2005-04-16 16:20:36 -06:00
|
|
|
fl->fl_break_time = break_time;
|
2011-07-26 16:25:49 -06:00
|
|
|
} else {
|
|
|
|
if (lease_breaking(flock))
|
|
|
|
continue;
|
|
|
|
fl->fl_flags |= FL_DOWNGRADE_PENDING;
|
|
|
|
fl->fl_downgrade_time = break_time;
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
2011-07-26 16:25:49 -06:00
|
|
|
fl->fl_lmops->lm_break(fl);
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
|
|
|
|
if (i_have_this_lease || (mode & O_NONBLOCK)) {
|
|
|
|
error = -EWOULDBLOCK;
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
restart:
|
|
|
|
break_time = flock->fl_break_time;
|
|
|
|
if (break_time != 0) {
|
|
|
|
break_time -= jiffies;
|
|
|
|
if (break_time == 0)
|
|
|
|
break_time++;
|
|
|
|
}
|
2008-01-14 21:28:30 -07:00
|
|
|
locks_insert_block(flock, new_fl);
|
2010-09-18 07:09:31 -06:00
|
|
|
unlock_flocks();
|
2008-01-14 21:28:30 -07:00
|
|
|
error = wait_event_interruptible_timeout(new_fl->fl_wait,
|
|
|
|
!new_fl->fl_next, break_time);
|
2010-09-18 07:09:31 -06:00
|
|
|
lock_flocks();
|
2008-01-14 21:28:30 -07:00
|
|
|
__locks_delete_block(new_fl);
|
2005-04-16 16:20:36 -06:00
|
|
|
if (error >= 0) {
|
|
|
|
if (error == 0)
|
|
|
|
time_out_leases(inode);
|
2011-07-26 16:25:49 -06:00
|
|
|
/*
|
|
|
|
* Wait for the next conflicting lease that has not been
|
|
|
|
* broken yet
|
|
|
|
*/
|
2005-04-16 16:20:36 -06:00
|
|
|
for (flock = inode->i_flock; flock && IS_LEASE(flock);
|
|
|
|
flock = flock->fl_next) {
|
2011-07-26 16:25:49 -06:00
|
|
|
if (locks_conflict(new_fl, flock))
|
2005-04-16 16:20:36 -06:00
|
|
|
goto restart;
|
|
|
|
}
|
|
|
|
error = 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
out:
|
2010-09-18 07:09:31 -06:00
|
|
|
unlock_flocks();
|
2011-12-26 11:25:26 -07:00
|
|
|
locks_free_lock(new_fl);
|
2005-04-16 16:20:36 -06:00
|
|
|
return error;
|
|
|
|
}
|
|
|
|
|
|
|
|
EXPORT_SYMBOL(__break_lease);
|
|
|
|
|
|
|
|
/**
|
2008-03-19 18:01:00 -06:00
|
|
|
* lease_get_mtime - get the last modified time of an inode
|
2005-04-16 16:20:36 -06:00
|
|
|
* @inode: the inode
|
|
|
|
* @time: pointer to a timespec which will contain the last modified time
|
|
|
|
*
|
|
|
|
* This is to force NFS clients to flush their caches for files with
|
|
|
|
* exclusive leases. The justification is that if someone has an
|
2008-03-19 18:01:00 -06:00
|
|
|
* exclusive lease, then they could be modifying it.
|
2005-04-16 16:20:36 -06:00
|
|
|
*/
|
|
|
|
void lease_get_mtime(struct inode *inode, struct timespec *time)
|
|
|
|
{
|
|
|
|
struct file_lock *flock = inode->i_flock;
|
2012-08-02 13:46:30 -06:00
|
|
|
if (flock && IS_LEASE(flock) && (flock->fl_type == F_WRLCK))
|
2005-04-16 16:20:36 -06:00
|
|
|
*time = current_fs_time(inode->i_sb);
|
|
|
|
else
|
|
|
|
*time = inode->i_mtime;
|
|
|
|
}
|
|
|
|
|
|
|
|
EXPORT_SYMBOL(lease_get_mtime);
|
|
|
|
|
|
|
|
/**
|
|
|
|
* fcntl_getlease - Enquire what lease is currently active
|
|
|
|
* @filp: the file
|
|
|
|
*
|
|
|
|
* The value returned by this function will be one of
|
|
|
|
* (if no lease break is pending):
|
|
|
|
*
|
|
|
|
* %F_RDLCK to indicate a shared lease is held.
|
|
|
|
*
|
|
|
|
* %F_WRLCK to indicate an exclusive lease is held.
|
|
|
|
*
|
|
|
|
* %F_UNLCK to indicate no lease is held.
|
|
|
|
*
|
|
|
|
* (if a lease break is pending):
|
|
|
|
*
|
|
|
|
* %F_RDLCK to indicate an exclusive lease needs to be
|
|
|
|
* changed to a shared lease (or removed).
|
|
|
|
*
|
|
|
|
* %F_UNLCK to indicate the lease needs to be removed.
|
|
|
|
*
|
|
|
|
* XXX: sfr & willy disagree over whether F_INPROGRESS
|
|
|
|
* should be returned to userspace.
|
|
|
|
*/
|
|
|
|
int fcntl_getlease(struct file *filp)
|
|
|
|
{
|
|
|
|
struct file_lock *fl;
|
|
|
|
int type = F_UNLCK;
|
|
|
|
|
2010-09-18 07:09:31 -06:00
|
|
|
lock_flocks();
|
2006-12-08 03:36:35 -07:00
|
|
|
time_out_leases(filp->f_path.dentry->d_inode);
|
|
|
|
for (fl = filp->f_path.dentry->d_inode->i_flock; fl && IS_LEASE(fl);
|
2005-04-16 16:20:36 -06:00
|
|
|
fl = fl->fl_next) {
|
|
|
|
if (fl->fl_file == filp) {
|
2011-07-26 16:25:49 -06:00
|
|
|
type = target_leasetype(fl);
|
2005-04-16 16:20:36 -06:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
2010-09-18 07:09:31 -06:00
|
|
|
unlock_flocks();
|
2005-04-16 16:20:36 -06:00
|
|
|
return type;
|
|
|
|
}
|
|
|
|
|
2011-09-21 06:34:32 -06:00
|
|
|
int generic_add_lease(struct file *filp, long arg, struct file_lock **flp)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
2005-07-07 18:57:06 -06:00
|
|
|
struct file_lock *fl, **before, **my_before = NULL, *lease;
|
2006-12-08 03:36:35 -07:00
|
|
|
struct dentry *dentry = filp->f_path.dentry;
|
2005-04-16 16:20:36 -06:00
|
|
|
struct inode *inode = dentry->d_inode;
|
2011-08-19 08:59:49 -06:00
|
|
|
int error;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2010-10-30 15:31:14 -06:00
|
|
|
lease = *flp;
|
|
|
|
|
2011-09-21 06:34:32 -06:00
|
|
|
error = -EAGAIN;
|
|
|
|
if ((arg == F_RDLCK) && (atomic_read(&inode->i_writecount) > 0))
|
2010-10-30 15:31:14 -06:00
|
|
|
goto out;
|
2011-09-21 06:34:32 -06:00
|
|
|
if ((arg == F_WRLCK)
|
|
|
|
&& ((dentry->d_count > 1)
|
|
|
|
|| (atomic_read(&inode->i_count) > 1)))
|
2010-10-30 15:31:14 -06:00
|
|
|
goto out;
|
2007-05-31 15:03:46 -06:00
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
/*
|
|
|
|
* At this point, we know that if there is an exclusive
|
|
|
|
* lease on this file, then we hold it on this filp
|
|
|
|
* (otherwise our open of this file would have blocked).
|
|
|
|
* And if we are trying to acquire an exclusive lease,
|
|
|
|
* then the file is not open by anyone (including us)
|
|
|
|
* except for this filp.
|
|
|
|
*/
|
2011-08-19 08:59:49 -06:00
|
|
|
error = -EAGAIN;
|
2005-04-16 16:20:36 -06:00
|
|
|
for (before = &inode->i_flock;
|
|
|
|
((fl = *before) != NULL) && IS_LEASE(fl);
|
|
|
|
before = &fl->fl_next) {
|
2011-08-19 08:59:49 -06:00
|
|
|
if (fl->fl_file == filp) {
|
2005-04-16 16:20:36 -06:00
|
|
|
my_before = before;
|
2011-08-19 08:59:49 -06:00
|
|
|
continue;
|
|
|
|
}
|
|
|
|
/*
|
|
|
|
* No exclusive leases if someone else has a lease on
|
|
|
|
* this file:
|
|
|
|
*/
|
|
|
|
if (arg == F_WRLCK)
|
|
|
|
goto out;
|
|
|
|
/*
|
|
|
|
* Modifying our existing lease is OK, but no getting a
|
|
|
|
* new lease if someone else is opening for write:
|
|
|
|
*/
|
|
|
|
if (fl->fl_flags & FL_UNLOCK_PENDING)
|
|
|
|
goto out;
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
|
|
|
|
if (my_before != NULL) {
|
2011-07-20 18:21:59 -06:00
|
|
|
error = lease->fl_lmops->lm_change(my_before, arg);
|
2010-10-31 06:35:10 -06:00
|
|
|
if (!error)
|
|
|
|
*flp = *my_before;
|
2005-04-16 16:20:36 -06:00
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
error = -EINVAL;
|
|
|
|
if (!leases_enable)
|
|
|
|
goto out;
|
|
|
|
|
2010-10-27 07:46:08 -06:00
|
|
|
locks_insert_lock(before, lease);
|
2007-09-20 02:45:02 -06:00
|
|
|
return 0;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
out:
|
|
|
|
return error;
|
|
|
|
}
|
2011-09-21 06:34:32 -06:00
|
|
|
|
|
|
|
int generic_delete_lease(struct file *filp, struct file_lock **flp)
|
|
|
|
{
|
|
|
|
struct file_lock *fl, **before;
|
|
|
|
struct dentry *dentry = filp->f_path.dentry;
|
|
|
|
struct inode *inode = dentry->d_inode;
|
|
|
|
|
|
|
|
for (before = &inode->i_flock;
|
|
|
|
((fl = *before) != NULL) && IS_LEASE(fl);
|
|
|
|
before = &fl->fl_next) {
|
|
|
|
if (fl->fl_file != filp)
|
|
|
|
continue;
|
|
|
|
return (*flp)->fl_lmops->lm_change(before, F_UNLCK);
|
|
|
|
}
|
|
|
|
return -EAGAIN;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* generic_setlease - sets a lease on an open file
|
|
|
|
* @filp: file pointer
|
|
|
|
* @arg: type of lease to obtain
|
|
|
|
* @flp: input - file_lock to use, output - file_lock inserted
|
|
|
|
*
|
|
|
|
* The (input) flp->fl_lmops->lm_break function is required
|
|
|
|
* by break_lease().
|
|
|
|
*
|
|
|
|
* Called with file_lock_lock held.
|
|
|
|
*/
|
|
|
|
int generic_setlease(struct file *filp, long arg, struct file_lock **flp)
|
|
|
|
{
|
|
|
|
struct dentry *dentry = filp->f_path.dentry;
|
|
|
|
struct inode *inode = dentry->d_inode;
|
|
|
|
int error;
|
|
|
|
|
2012-03-03 22:17:15 -07:00
|
|
|
if ((!uid_eq(current_fsuid(), inode->i_uid)) && !capable(CAP_LEASE))
|
2011-09-21 06:34:32 -06:00
|
|
|
return -EACCES;
|
|
|
|
if (!S_ISREG(inode->i_mode))
|
|
|
|
return -EINVAL;
|
|
|
|
error = security_file_lock(filp, arg);
|
|
|
|
if (error)
|
|
|
|
return error;
|
|
|
|
|
|
|
|
time_out_leases(inode);
|
|
|
|
|
|
|
|
BUG_ON(!(*flp)->fl_lmops->lm_break);
|
|
|
|
|
|
|
|
switch (arg) {
|
|
|
|
case F_UNLCK:
|
|
|
|
return generic_delete_lease(filp, flp);
|
|
|
|
case F_RDLCK:
|
|
|
|
case F_WRLCK:
|
|
|
|
return generic_add_lease(filp, arg, flp);
|
|
|
|
default:
|
2012-07-13 11:35:36 -06:00
|
|
|
return -EINVAL;
|
2011-09-21 06:34:32 -06:00
|
|
|
}
|
|
|
|
}
|
2007-07-31 01:39:22 -06:00
|
|
|
EXPORT_SYMBOL(generic_setlease);
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2010-09-18 07:09:31 -06:00
|
|
|
static int __vfs_setlease(struct file *filp, long arg, struct file_lock **lease)
|
|
|
|
{
|
|
|
|
if (filp->f_op && filp->f_op->setlease)
|
|
|
|
return filp->f_op->setlease(filp, arg, lease);
|
|
|
|
else
|
|
|
|
return generic_setlease(filp, arg, lease);
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
2007-06-07 15:09:49 -06:00
|
|
|
* vfs_setlease - sets a lease on an open file
|
2005-04-16 16:20:36 -06:00
|
|
|
* @filp: file pointer
|
|
|
|
* @arg: type of lease to obtain
|
|
|
|
* @lease: file_lock to use
|
|
|
|
*
|
|
|
|
* Call this to establish a lease on the file.
|
2011-07-20 18:21:59 -06:00
|
|
|
* The (*lease)->fl_lmops->lm_break operation must be set; if not,
|
2006-11-14 13:51:40 -07:00
|
|
|
* break_lease will oops!
|
|
|
|
*
|
|
|
|
* This will call the filesystem's setlease file method, if
|
|
|
|
* defined. Note that there is no getlease method; instead, the
|
|
|
|
* filesystem setlease method should call back to setlease() to
|
|
|
|
* add a lease to the inode's lease list, where fcntl_getlease() can
|
|
|
|
* find it. Since fcntl_getlease() only reports whether the current
|
|
|
|
* task holds a lease, a cluster filesystem need only do this for
|
|
|
|
* leases held by processes on this node.
|
|
|
|
*
|
|
|
|
* There is also no break_lease method; filesystems that
|
2009-12-18 13:40:42 -07:00
|
|
|
* handle their own leases should break leases themselves from the
|
2006-11-14 13:51:40 -07:00
|
|
|
* filesystem's open, create, and (on truncate) setattr methods.
|
|
|
|
*
|
|
|
|
* Warning: the only current setlease methods exist only to disable
|
|
|
|
* leases in certain cases. More vfs changes may be required to
|
|
|
|
* allow a full filesystem lease implementation.
|
2005-04-16 16:20:36 -06:00
|
|
|
*/
|
|
|
|
|
2007-06-07 15:09:49 -06:00
|
|
|
int vfs_setlease(struct file *filp, long arg, struct file_lock **lease)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
|
|
|
int error;
|
|
|
|
|
2010-09-18 07:09:31 -06:00
|
|
|
lock_flocks();
|
|
|
|
error = __vfs_setlease(filp, arg, lease);
|
|
|
|
unlock_flocks();
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
return error;
|
|
|
|
}
|
2007-06-07 15:09:49 -06:00
|
|
|
EXPORT_SYMBOL_GPL(vfs_setlease);
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2010-10-30 15:31:13 -06:00
|
|
|
static int do_fcntl_delete_lease(struct file *filp)
|
|
|
|
{
|
|
|
|
struct file_lock fl, *flp = &fl;
|
|
|
|
|
|
|
|
lease_init(filp, F_UNLCK, flp);
|
|
|
|
|
|
|
|
return vfs_setlease(filp, F_UNLCK, &flp);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int do_fcntl_add_lease(unsigned int fd, struct file *filp, long arg)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
2010-11-03 14:49:44 -06:00
|
|
|
struct file_lock *fl, *ret;
|
2010-10-27 10:38:12 -06:00
|
|
|
struct fasync_struct *new;
|
2005-04-16 16:20:36 -06:00
|
|
|
int error;
|
|
|
|
|
2010-10-27 07:46:08 -06:00
|
|
|
fl = lease_alloc(filp, arg);
|
|
|
|
if (IS_ERR(fl))
|
|
|
|
return PTR_ERR(fl);
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2010-10-27 10:38:12 -06:00
|
|
|
new = fasync_alloc();
|
|
|
|
if (!new) {
|
|
|
|
locks_free_lock(fl);
|
|
|
|
return -ENOMEM;
|
|
|
|
}
|
2010-11-03 14:49:44 -06:00
|
|
|
ret = fl;
|
2010-09-18 07:09:31 -06:00
|
|
|
lock_flocks();
|
2010-11-03 16:09:18 -06:00
|
|
|
error = __vfs_setlease(filp, arg, &ret);
|
2010-10-31 06:35:10 -06:00
|
|
|
if (error) {
|
|
|
|
unlock_flocks();
|
|
|
|
locks_free_lock(fl);
|
|
|
|
goto out_free_fasync;
|
|
|
|
}
|
2010-11-03 14:49:44 -06:00
|
|
|
if (ret != fl)
|
|
|
|
locks_free_lock(fl);
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2010-10-27 10:38:12 -06:00
|
|
|
/*
|
|
|
|
* fasync_insert_entry() returns the old entry if any.
|
|
|
|
* If there was no old entry, then it used 'new' and
|
|
|
|
* inserted it into the fasync list. Clear new so that
|
|
|
|
* we don't release it here.
|
|
|
|
*/
|
2010-11-03 14:49:44 -06:00
|
|
|
if (!fasync_insert_entry(fd, filp, &ret->fl_fasync, new))
|
2010-10-27 10:38:12 -06:00
|
|
|
new = NULL;
|
|
|
|
|
2010-11-03 16:09:18 -06:00
|
|
|
error = __f_setown(filp, task_pid(current), PIDTYPE_PID, 0);
|
2010-09-18 07:09:31 -06:00
|
|
|
unlock_flocks();
|
2010-10-31 06:35:10 -06:00
|
|
|
|
|
|
|
out_free_fasync:
|
2010-10-27 10:38:12 -06:00
|
|
|
if (new)
|
|
|
|
fasync_free(new);
|
2005-04-16 16:20:36 -06:00
|
|
|
return error;
|
|
|
|
}
|
|
|
|
|
2010-10-30 15:31:13 -06:00
|
|
|
/**
|
|
|
|
* fcntl_setlease - sets a lease on an open file
|
|
|
|
* @fd: open file descriptor
|
|
|
|
* @filp: file pointer
|
|
|
|
* @arg: type of lease to obtain
|
|
|
|
*
|
|
|
|
* Call this fcntl to establish a lease on the file.
|
|
|
|
* Note that you also need to call %F_SETSIG to
|
|
|
|
* receive a signal when the lease is broken.
|
|
|
|
*/
|
|
|
|
int fcntl_setlease(unsigned int fd, struct file *filp, long arg)
|
|
|
|
{
|
|
|
|
if (arg == F_UNLCK)
|
|
|
|
return do_fcntl_delete_lease(filp);
|
|
|
|
return do_fcntl_add_lease(fd, filp, arg);
|
|
|
|
}
|
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
/**
|
|
|
|
* flock_lock_file_wait - Apply a FLOCK-style lock to a file
|
|
|
|
* @filp: The file to apply the lock to
|
|
|
|
* @fl: The lock to be applied
|
|
|
|
*
|
|
|
|
* Add a FLOCK style lock to a file.
|
|
|
|
*/
|
|
|
|
int flock_lock_file_wait(struct file *filp, struct file_lock *fl)
|
|
|
|
{
|
|
|
|
int error;
|
|
|
|
might_sleep();
|
|
|
|
for (;;) {
|
|
|
|
error = flock_lock_file(filp, fl);
|
2008-07-25 02:48:57 -06:00
|
|
|
if (error != FILE_LOCK_DEFERRED)
|
2005-04-16 16:20:36 -06:00
|
|
|
break;
|
|
|
|
error = wait_event_interruptible(fl->fl_wait, !fl->fl_next);
|
|
|
|
if (!error)
|
|
|
|
continue;
|
|
|
|
|
|
|
|
locks_delete_block(fl);
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
return error;
|
|
|
|
}
|
|
|
|
|
|
|
|
EXPORT_SYMBOL(flock_lock_file_wait);
|
|
|
|
|
|
|
|
/**
|
|
|
|
* sys_flock: - flock() system call.
|
|
|
|
* @fd: the file descriptor to lock.
|
|
|
|
* @cmd: the type of lock to apply.
|
|
|
|
*
|
|
|
|
* Apply a %FL_FLOCK style lock to an open file descriptor.
|
|
|
|
* The @cmd can be one of
|
|
|
|
*
|
|
|
|
* %LOCK_SH -- a shared lock.
|
|
|
|
*
|
|
|
|
* %LOCK_EX -- an exclusive lock.
|
|
|
|
*
|
|
|
|
* %LOCK_UN -- remove an existing lock.
|
|
|
|
*
|
|
|
|
* %LOCK_MAND -- a `mandatory' flock. This exists to emulate Windows Share Modes.
|
|
|
|
*
|
|
|
|
* %LOCK_MAND can be combined with %LOCK_READ or %LOCK_WRITE to allow other
|
|
|
|
* processes read and write access respectively.
|
|
|
|
*/
|
2009-01-14 06:14:18 -07:00
|
|
|
SYSCALL_DEFINE2(flock, unsigned int, fd, unsigned int, cmd)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
|
|
|
struct file *filp;
|
2012-04-21 16:46:53 -06:00
|
|
|
int fput_needed;
|
2005-04-16 16:20:36 -06:00
|
|
|
struct file_lock *lock;
|
|
|
|
int can_sleep, unlock;
|
|
|
|
int error;
|
|
|
|
|
|
|
|
error = -EBADF;
|
2012-04-21 16:46:53 -06:00
|
|
|
filp = fget_light(fd, &fput_needed);
|
2005-04-16 16:20:36 -06:00
|
|
|
if (!filp)
|
|
|
|
goto out;
|
|
|
|
|
|
|
|
can_sleep = !(cmd & LOCK_NB);
|
|
|
|
cmd &= ~LOCK_NB;
|
|
|
|
unlock = (cmd == LOCK_UN);
|
|
|
|
|
2008-09-02 13:28:45 -06:00
|
|
|
if (!unlock && !(cmd & LOCK_MAND) &&
|
|
|
|
!(filp->f_mode & (FMODE_READ|FMODE_WRITE)))
|
2005-04-16 16:20:36 -06:00
|
|
|
goto out_putf;
|
|
|
|
|
|
|
|
error = flock_make_lock(filp, &lock, cmd);
|
|
|
|
if (error)
|
|
|
|
goto out_putf;
|
|
|
|
if (can_sleep)
|
|
|
|
lock->fl_flags |= FL_SLEEP;
|
|
|
|
|
2009-07-16 01:41:39 -06:00
|
|
|
error = security_file_lock(filp, lock->fl_type);
|
2005-04-16 16:20:36 -06:00
|
|
|
if (error)
|
|
|
|
goto out_free;
|
|
|
|
|
|
|
|
if (filp->f_op && filp->f_op->flock)
|
|
|
|
error = filp->f_op->flock(filp,
|
|
|
|
(can_sleep) ? F_SETLKW : F_SETLK,
|
|
|
|
lock);
|
|
|
|
else
|
|
|
|
error = flock_lock_file_wait(filp, lock);
|
|
|
|
|
|
|
|
out_free:
|
2006-03-31 03:30:55 -07:00
|
|
|
locks_free_lock(lock);
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
out_putf:
|
2012-04-21 16:46:53 -06:00
|
|
|
fput_light(filp, fput_needed);
|
2005-04-16 16:20:36 -06:00
|
|
|
out:
|
|
|
|
return error;
|
|
|
|
}
|
|
|
|
|
2007-02-20 22:58:50 -07:00
|
|
|
/**
|
|
|
|
* vfs_test_lock - test file byte range lock
|
|
|
|
* @filp: The file to test lock for
|
2007-05-11 14:22:50 -06:00
|
|
|
* @fl: The lock to test; also used to hold result
|
2007-02-20 22:58:50 -07:00
|
|
|
*
|
|
|
|
* Returns -ERRNO on failure. Indicates presence of conflicting lock by
|
|
|
|
* setting conf->fl_type to something other than F_UNLCK.
|
|
|
|
*/
|
|
|
|
int vfs_test_lock(struct file *filp, struct file_lock *fl)
|
|
|
|
{
|
|
|
|
if (filp->f_op && filp->f_op->lock)
|
|
|
|
return filp->f_op->lock(filp, F_GETLK, fl);
|
|
|
|
posix_test_lock(filp, fl);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(vfs_test_lock);
|
|
|
|
|
2007-02-20 14:10:11 -07:00
|
|
|
static int posix_lock_to_flock(struct flock *flock, struct file_lock *fl)
|
|
|
|
{
|
|
|
|
flock->l_pid = fl->fl_pid;
|
|
|
|
#if BITS_PER_LONG == 32
|
|
|
|
/*
|
|
|
|
* Make sure we can represent the posix lock via
|
|
|
|
* legacy 32bit flock.
|
|
|
|
*/
|
|
|
|
if (fl->fl_start > OFFT_OFFSET_MAX)
|
|
|
|
return -EOVERFLOW;
|
|
|
|
if (fl->fl_end != OFFSET_MAX && fl->fl_end > OFFT_OFFSET_MAX)
|
|
|
|
return -EOVERFLOW;
|
|
|
|
#endif
|
|
|
|
flock->l_start = fl->fl_start;
|
|
|
|
flock->l_len = fl->fl_end == OFFSET_MAX ? 0 :
|
|
|
|
fl->fl_end - fl->fl_start + 1;
|
|
|
|
flock->l_whence = 0;
|
2007-05-10 16:38:43 -06:00
|
|
|
flock->l_type = fl->fl_type;
|
2007-02-20 14:10:11 -07:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
#if BITS_PER_LONG == 32
|
|
|
|
static void posix_lock_to_flock64(struct flock64 *flock, struct file_lock *fl)
|
|
|
|
{
|
|
|
|
flock->l_pid = fl->fl_pid;
|
|
|
|
flock->l_start = fl->fl_start;
|
|
|
|
flock->l_len = fl->fl_end == OFFSET_MAX ? 0 :
|
|
|
|
fl->fl_end - fl->fl_start + 1;
|
|
|
|
flock->l_whence = 0;
|
|
|
|
flock->l_type = fl->fl_type;
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
/* Report the first existing lock that would conflict with l.
|
|
|
|
* This implements the F_GETLK command of fcntl().
|
|
|
|
*/
|
|
|
|
int fcntl_getlk(struct file *filp, struct flock __user *l)
|
|
|
|
{
|
2007-02-20 22:55:18 -07:00
|
|
|
struct file_lock file_lock;
|
2005-04-16 16:20:36 -06:00
|
|
|
struct flock flock;
|
|
|
|
int error;
|
|
|
|
|
|
|
|
error = -EFAULT;
|
|
|
|
if (copy_from_user(&flock, l, sizeof(flock)))
|
|
|
|
goto out;
|
|
|
|
error = -EINVAL;
|
|
|
|
if ((flock.l_type != F_RDLCK) && (flock.l_type != F_WRLCK))
|
|
|
|
goto out;
|
|
|
|
|
|
|
|
error = flock_to_posix_lock(filp, &file_lock, &flock);
|
|
|
|
if (error)
|
|
|
|
goto out;
|
|
|
|
|
2007-02-20 22:58:50 -07:00
|
|
|
error = vfs_test_lock(filp, &file_lock);
|
|
|
|
if (error)
|
|
|
|
goto out;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2007-02-20 22:55:18 -07:00
|
|
|
flock.l_type = file_lock.fl_type;
|
|
|
|
if (file_lock.fl_type != F_UNLCK) {
|
|
|
|
error = posix_lock_to_flock(&flock, &file_lock);
|
2007-02-20 14:10:11 -07:00
|
|
|
if (error)
|
2005-04-16 16:20:36 -06:00
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
error = -EFAULT;
|
|
|
|
if (!copy_to_user(l, &flock, sizeof(flock)))
|
|
|
|
error = 0;
|
|
|
|
out:
|
|
|
|
return error;
|
|
|
|
}
|
|
|
|
|
2007-01-18 13:08:55 -07:00
|
|
|
/**
|
|
|
|
* vfs_lock_file - file byte range lock
|
|
|
|
* @filp: The file to apply the lock to
|
|
|
|
* @cmd: type of locking operation (F_SETLK, F_GETLK, etc.)
|
|
|
|
* @fl: The lock to be applied
|
2007-01-18 14:15:35 -07:00
|
|
|
* @conf: Place to return a copy of the conflicting lock, if found.
|
|
|
|
*
|
|
|
|
* A caller that doesn't care about the conflicting lock may pass NULL
|
|
|
|
* as the final argument.
|
|
|
|
*
|
|
|
|
* If the filesystem defines a private ->lock() method, then @conf will
|
|
|
|
* be left unchanged; so a caller that cares should initialize it to
|
|
|
|
* some acceptable default.
|
locks: add fl_grant callback for asynchronous lock return
Acquiring a lock on a cluster filesystem may require communication with
remote hosts, and to avoid blocking lockd or nfsd threads during such
communication, we allow the results to be returned asynchronously.
When a ->lock() call needs to block, the file system will return
-EINPROGRESS, and then later return the results with a call to the
routine in the fl_grant field of the lock_manager_operations struct.
This differs from the case when ->lock returns -EAGAIN to a blocking
lock request; in that case, the filesystem calls fl_notify when the lock
is granted, and the caller retries the original lock. So while
fl_notify is merely a hint to the caller that it should retry, fl_grant
actually communicates the final result of the lock operation (with the
lock already acquired in the succesful case).
Therefore fl_grant takes a lock, a status and, for the test lock case, a
conflicting lock. We also allow fl_grant to return an error to the
filesystem, to handle the case where the fl_grant requests arrives after
the lock manager has already given up waiting for it.
Signed-off-by: Marc Eshel <eshel@almaden.ibm.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
2006-12-05 21:31:28 -07:00
|
|
|
*
|
|
|
|
* To avoid blocking kernel daemons, such as lockd, that need to acquire POSIX
|
|
|
|
* locks, the ->lock() interface may return asynchronously, before the lock has
|
|
|
|
* been granted or denied by the underlying filesystem, if (and only if)
|
2011-07-20 18:21:59 -06:00
|
|
|
* lm_grant is set. Callers expecting ->lock() to return asynchronously
|
locks: add fl_grant callback for asynchronous lock return
Acquiring a lock on a cluster filesystem may require communication with
remote hosts, and to avoid blocking lockd or nfsd threads during such
communication, we allow the results to be returned asynchronously.
When a ->lock() call needs to block, the file system will return
-EINPROGRESS, and then later return the results with a call to the
routine in the fl_grant field of the lock_manager_operations struct.
This differs from the case when ->lock returns -EAGAIN to a blocking
lock request; in that case, the filesystem calls fl_notify when the lock
is granted, and the caller retries the original lock. So while
fl_notify is merely a hint to the caller that it should retry, fl_grant
actually communicates the final result of the lock operation (with the
lock already acquired in the succesful case).
Therefore fl_grant takes a lock, a status and, for the test lock case, a
conflicting lock. We also allow fl_grant to return an error to the
filesystem, to handle the case where the fl_grant requests arrives after
the lock manager has already given up waiting for it.
Signed-off-by: Marc Eshel <eshel@almaden.ibm.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
2006-12-05 21:31:28 -07:00
|
|
|
* will only use F_SETLK, not F_SETLKW; they will set FL_SLEEP if (and only if)
|
|
|
|
* the request is for a blocking lock. When ->lock() does return asynchronously,
|
2011-07-20 18:21:59 -06:00
|
|
|
* it must return FILE_LOCK_DEFERRED, and call ->lm_grant() when the lock
|
locks: add fl_grant callback for asynchronous lock return
Acquiring a lock on a cluster filesystem may require communication with
remote hosts, and to avoid blocking lockd or nfsd threads during such
communication, we allow the results to be returned asynchronously.
When a ->lock() call needs to block, the file system will return
-EINPROGRESS, and then later return the results with a call to the
routine in the fl_grant field of the lock_manager_operations struct.
This differs from the case when ->lock returns -EAGAIN to a blocking
lock request; in that case, the filesystem calls fl_notify when the lock
is granted, and the caller retries the original lock. So while
fl_notify is merely a hint to the caller that it should retry, fl_grant
actually communicates the final result of the lock operation (with the
lock already acquired in the succesful case).
Therefore fl_grant takes a lock, a status and, for the test lock case, a
conflicting lock. We also allow fl_grant to return an error to the
filesystem, to handle the case where the fl_grant requests arrives after
the lock manager has already given up waiting for it.
Signed-off-by: Marc Eshel <eshel@almaden.ibm.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
2006-12-05 21:31:28 -07:00
|
|
|
* request completes.
|
|
|
|
* If the request is for non-blocking lock the file system should return
|
2008-07-25 02:48:57 -06:00
|
|
|
* FILE_LOCK_DEFERRED then try to get the lock and call the callback routine
|
|
|
|
* with the result. If the request timed out the callback routine will return a
|
locks: add fl_grant callback for asynchronous lock return
Acquiring a lock on a cluster filesystem may require communication with
remote hosts, and to avoid blocking lockd or nfsd threads during such
communication, we allow the results to be returned asynchronously.
When a ->lock() call needs to block, the file system will return
-EINPROGRESS, and then later return the results with a call to the
routine in the fl_grant field of the lock_manager_operations struct.
This differs from the case when ->lock returns -EAGAIN to a blocking
lock request; in that case, the filesystem calls fl_notify when the lock
is granted, and the caller retries the original lock. So while
fl_notify is merely a hint to the caller that it should retry, fl_grant
actually communicates the final result of the lock operation (with the
lock already acquired in the succesful case).
Therefore fl_grant takes a lock, a status and, for the test lock case, a
conflicting lock. We also allow fl_grant to return an error to the
filesystem, to handle the case where the fl_grant requests arrives after
the lock manager has already given up waiting for it.
Signed-off-by: Marc Eshel <eshel@almaden.ibm.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
2006-12-05 21:31:28 -07:00
|
|
|
* nonzero return code and the file system should release the lock. The file
|
|
|
|
* system is also responsible to keep a corresponding posix lock when it
|
|
|
|
* grants a lock so the VFS can find out which locks are locally held and do
|
|
|
|
* the correct lock cleanup when required.
|
|
|
|
* The underlying filesystem must not drop the kernel lock or call
|
2011-07-20 18:21:59 -06:00
|
|
|
* ->lm_grant() before returning to the caller with a FILE_LOCK_DEFERRED
|
locks: add fl_grant callback for asynchronous lock return
Acquiring a lock on a cluster filesystem may require communication with
remote hosts, and to avoid blocking lockd or nfsd threads during such
communication, we allow the results to be returned asynchronously.
When a ->lock() call needs to block, the file system will return
-EINPROGRESS, and then later return the results with a call to the
routine in the fl_grant field of the lock_manager_operations struct.
This differs from the case when ->lock returns -EAGAIN to a blocking
lock request; in that case, the filesystem calls fl_notify when the lock
is granted, and the caller retries the original lock. So while
fl_notify is merely a hint to the caller that it should retry, fl_grant
actually communicates the final result of the lock operation (with the
lock already acquired in the succesful case).
Therefore fl_grant takes a lock, a status and, for the test lock case, a
conflicting lock. We also allow fl_grant to return an error to the
filesystem, to handle the case where the fl_grant requests arrives after
the lock manager has already given up waiting for it.
Signed-off-by: Marc Eshel <eshel@almaden.ibm.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
2006-12-05 21:31:28 -07:00
|
|
|
* return code.
|
2007-01-18 13:08:55 -07:00
|
|
|
*/
|
2007-01-18 14:15:35 -07:00
|
|
|
int vfs_lock_file(struct file *filp, unsigned int cmd, struct file_lock *fl, struct file_lock *conf)
|
2007-01-18 13:08:55 -07:00
|
|
|
{
|
|
|
|
if (filp->f_op && filp->f_op->lock)
|
|
|
|
return filp->f_op->lock(filp, cmd, fl);
|
|
|
|
else
|
2007-01-18 14:15:35 -07:00
|
|
|
return posix_lock_file(filp, fl, conf);
|
2007-01-18 13:08:55 -07:00
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(vfs_lock_file);
|
|
|
|
|
2008-07-25 02:48:57 -06:00
|
|
|
static int do_lock_file_wait(struct file *filp, unsigned int cmd,
|
|
|
|
struct file_lock *fl)
|
|
|
|
{
|
|
|
|
int error;
|
|
|
|
|
|
|
|
error = security_file_lock(filp, fl->fl_type);
|
|
|
|
if (error)
|
|
|
|
return error;
|
|
|
|
|
2008-07-25 02:48:58 -06:00
|
|
|
for (;;) {
|
|
|
|
error = vfs_lock_file(filp, cmd, fl, NULL);
|
|
|
|
if (error != FILE_LOCK_DEFERRED)
|
2008-07-25 02:48:57 -06:00
|
|
|
break;
|
2008-07-25 02:48:58 -06:00
|
|
|
error = wait_event_interruptible(fl->fl_wait, !fl->fl_next);
|
|
|
|
if (!error)
|
|
|
|
continue;
|
|
|
|
|
|
|
|
locks_delete_block(fl);
|
|
|
|
break;
|
2008-07-25 02:48:57 -06:00
|
|
|
}
|
|
|
|
|
|
|
|
return error;
|
|
|
|
}
|
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
/* Apply the lock described by l to an open file descriptor.
|
|
|
|
* This implements both the F_SETLK and F_SETLKW commands of fcntl().
|
|
|
|
*/
|
[PATCH] stale POSIX lock handling
I believe that there is a problem with the handling of POSIX locks, which
the attached patch should address.
The problem appears to be a race between fcntl(2) and close(2). A
multithreaded application could close a file descriptor at the same time as
it is trying to acquire a lock using the same file descriptor. I would
suggest that that multithreaded application is not providing the proper
synchronization for itself, but the OS should still behave correctly.
SUS3 (Single UNIX Specification Version 3, read: POSIX) indicates that when
a file descriptor is closed, that all POSIX locks on the file, owned by the
process which closed the file descriptor, should be released.
The trick here is when those locks are released. The current code releases
all locks which exist when close is processing, but any locks in progress
are handled when the last reference to the open file is released.
There are three cases to consider.
One is the simple case, a multithreaded (mt) process has a file open and
races to close it and acquire a lock on it. In this case, the close will
release one reference to the open file and when the fcntl is done, it will
release the other reference. For this situation, no locks should exist on
the file when both the close and fcntl operations are done. The current
system will handle this case because the last reference to the open file is
being released.
The second case is when the mt process has dup(2)'d the file descriptor.
The close will release one reference to the file and the fcntl, when done,
will release another, but there will still be at least one more reference
to the open file. One could argue that the existence of a lock on the file
after the close has completed is okay, because it was acquired after the
close operation and there is still a way for the application to release the
lock on the file, using an existing file descriptor.
The third case is when the mt process has forked, after opening the file
and either before or after becoming an mt process. In this case, each
process would hold a reference to the open file. For each process, this
degenerates to first case above. However, the lock continues to exist
until both processes have released their references to the open file. This
lock could block other lock requests.
The changes to release the lock when the last reference to the open file
aren't quite right because they would allow the lock to exist as long as
there was a reference to the open file. This is too long.
The new proposed solution is to add support in the fcntl code path to
detect a race with close and then to release the lock which was just
acquired when such as race is detected. This causes locks to be released
in a timely fashion and for the system to conform to the POSIX semantic
specification.
This was tested by instrumenting a kernel to detect the handling locks and
then running a program which generates case #3 above. A dangling lock
could be reliably generated. When the changes to detect the close/fcntl
race were added, a dangling lock could no longer be generated.
Cc: Matthew Wilcox <willy@debian.org>
Cc: Trond Myklebust <trond.myklebust@fys.uio.no>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-07-27 12:45:09 -06:00
|
|
|
int fcntl_setlk(unsigned int fd, struct file *filp, unsigned int cmd,
|
|
|
|
struct flock __user *l)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
|
|
|
struct file_lock *file_lock = locks_alloc_lock();
|
|
|
|
struct flock flock;
|
|
|
|
struct inode *inode;
|
2008-05-06 11:58:34 -06:00
|
|
|
struct file *f;
|
2005-04-16 16:20:36 -06:00
|
|
|
int error;
|
|
|
|
|
|
|
|
if (file_lock == NULL)
|
|
|
|
return -ENOLCK;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* This might block, so we do it before checking the inode.
|
|
|
|
*/
|
|
|
|
error = -EFAULT;
|
|
|
|
if (copy_from_user(&flock, l, sizeof(flock)))
|
|
|
|
goto out;
|
|
|
|
|
2006-12-08 03:36:35 -07:00
|
|
|
inode = filp->f_path.dentry->d_inode;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
/* Don't allow mandatory locks on files that may be memory mapped
|
|
|
|
* and shared.
|
|
|
|
*/
|
2007-10-01 15:41:11 -06:00
|
|
|
if (mandatory_lock(inode) && mapping_writably_mapped(filp->f_mapping)) {
|
2005-04-16 16:20:36 -06:00
|
|
|
error = -EAGAIN;
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
[PATCH] stale POSIX lock handling
I believe that there is a problem with the handling of POSIX locks, which
the attached patch should address.
The problem appears to be a race between fcntl(2) and close(2). A
multithreaded application could close a file descriptor at the same time as
it is trying to acquire a lock using the same file descriptor. I would
suggest that that multithreaded application is not providing the proper
synchronization for itself, but the OS should still behave correctly.
SUS3 (Single UNIX Specification Version 3, read: POSIX) indicates that when
a file descriptor is closed, that all POSIX locks on the file, owned by the
process which closed the file descriptor, should be released.
The trick here is when those locks are released. The current code releases
all locks which exist when close is processing, but any locks in progress
are handled when the last reference to the open file is released.
There are three cases to consider.
One is the simple case, a multithreaded (mt) process has a file open and
races to close it and acquire a lock on it. In this case, the close will
release one reference to the open file and when the fcntl is done, it will
release the other reference. For this situation, no locks should exist on
the file when both the close and fcntl operations are done. The current
system will handle this case because the last reference to the open file is
being released.
The second case is when the mt process has dup(2)'d the file descriptor.
The close will release one reference to the file and the fcntl, when done,
will release another, but there will still be at least one more reference
to the open file. One could argue that the existence of a lock on the file
after the close has completed is okay, because it was acquired after the
close operation and there is still a way for the application to release the
lock on the file, using an existing file descriptor.
The third case is when the mt process has forked, after opening the file
and either before or after becoming an mt process. In this case, each
process would hold a reference to the open file. For each process, this
degenerates to first case above. However, the lock continues to exist
until both processes have released their references to the open file. This
lock could block other lock requests.
The changes to release the lock when the last reference to the open file
aren't quite right because they would allow the lock to exist as long as
there was a reference to the open file. This is too long.
The new proposed solution is to add support in the fcntl code path to
detect a race with close and then to release the lock which was just
acquired when such as race is detected. This causes locks to be released
in a timely fashion and for the system to conform to the POSIX semantic
specification.
This was tested by instrumenting a kernel to detect the handling locks and
then running a program which generates case #3 above. A dangling lock
could be reliably generated. When the changes to detect the close/fcntl
race were added, a dangling lock could no longer be generated.
Cc: Matthew Wilcox <willy@debian.org>
Cc: Trond Myklebust <trond.myklebust@fys.uio.no>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-07-27 12:45:09 -06:00
|
|
|
again:
|
2005-04-16 16:20:36 -06:00
|
|
|
error = flock_to_posix_lock(filp, file_lock, &flock);
|
|
|
|
if (error)
|
|
|
|
goto out;
|
|
|
|
if (cmd == F_SETLKW) {
|
|
|
|
file_lock->fl_flags |= FL_SLEEP;
|
|
|
|
}
|
|
|
|
|
|
|
|
error = -EBADF;
|
|
|
|
switch (flock.l_type) {
|
|
|
|
case F_RDLCK:
|
|
|
|
if (!(filp->f_mode & FMODE_READ))
|
|
|
|
goto out;
|
|
|
|
break;
|
|
|
|
case F_WRLCK:
|
|
|
|
if (!(filp->f_mode & FMODE_WRITE))
|
|
|
|
goto out;
|
|
|
|
break;
|
|
|
|
case F_UNLCK:
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
error = -EINVAL;
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
2008-07-25 02:48:57 -06:00
|
|
|
error = do_lock_file_wait(filp, cmd, file_lock);
|
2005-04-16 16:20:36 -06:00
|
|
|
|
[PATCH] stale POSIX lock handling
I believe that there is a problem with the handling of POSIX locks, which
the attached patch should address.
The problem appears to be a race between fcntl(2) and close(2). A
multithreaded application could close a file descriptor at the same time as
it is trying to acquire a lock using the same file descriptor. I would
suggest that that multithreaded application is not providing the proper
synchronization for itself, but the OS should still behave correctly.
SUS3 (Single UNIX Specification Version 3, read: POSIX) indicates that when
a file descriptor is closed, that all POSIX locks on the file, owned by the
process which closed the file descriptor, should be released.
The trick here is when those locks are released. The current code releases
all locks which exist when close is processing, but any locks in progress
are handled when the last reference to the open file is released.
There are three cases to consider.
One is the simple case, a multithreaded (mt) process has a file open and
races to close it and acquire a lock on it. In this case, the close will
release one reference to the open file and when the fcntl is done, it will
release the other reference. For this situation, no locks should exist on
the file when both the close and fcntl operations are done. The current
system will handle this case because the last reference to the open file is
being released.
The second case is when the mt process has dup(2)'d the file descriptor.
The close will release one reference to the file and the fcntl, when done,
will release another, but there will still be at least one more reference
to the open file. One could argue that the existence of a lock on the file
after the close has completed is okay, because it was acquired after the
close operation and there is still a way for the application to release the
lock on the file, using an existing file descriptor.
The third case is when the mt process has forked, after opening the file
and either before or after becoming an mt process. In this case, each
process would hold a reference to the open file. For each process, this
degenerates to first case above. However, the lock continues to exist
until both processes have released their references to the open file. This
lock could block other lock requests.
The changes to release the lock when the last reference to the open file
aren't quite right because they would allow the lock to exist as long as
there was a reference to the open file. This is too long.
The new proposed solution is to add support in the fcntl code path to
detect a race with close and then to release the lock which was just
acquired when such as race is detected. This causes locks to be released
in a timely fashion and for the system to conform to the POSIX semantic
specification.
This was tested by instrumenting a kernel to detect the handling locks and
then running a program which generates case #3 above. A dangling lock
could be reliably generated. When the changes to detect the close/fcntl
race were added, a dangling lock could no longer be generated.
Cc: Matthew Wilcox <willy@debian.org>
Cc: Trond Myklebust <trond.myklebust@fys.uio.no>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-07-27 12:45:09 -06:00
|
|
|
/*
|
|
|
|
* Attempt to detect a close/fcntl race and recover by
|
|
|
|
* releasing the lock that was just acquired.
|
|
|
|
*/
|
2008-05-06 11:58:34 -06:00
|
|
|
/*
|
|
|
|
* we need that spin_lock here - it prevents reordering between
|
|
|
|
* update of inode->i_flock and check for it done in close().
|
|
|
|
* rcu_read_lock() wouldn't do.
|
|
|
|
*/
|
|
|
|
spin_lock(¤t->files->file_lock);
|
|
|
|
f = fcheck(fd);
|
|
|
|
spin_unlock(¤t->files->file_lock);
|
|
|
|
if (!error && f != filp && flock.l_type != F_UNLCK) {
|
[PATCH] stale POSIX lock handling
I believe that there is a problem with the handling of POSIX locks, which
the attached patch should address.
The problem appears to be a race between fcntl(2) and close(2). A
multithreaded application could close a file descriptor at the same time as
it is trying to acquire a lock using the same file descriptor. I would
suggest that that multithreaded application is not providing the proper
synchronization for itself, but the OS should still behave correctly.
SUS3 (Single UNIX Specification Version 3, read: POSIX) indicates that when
a file descriptor is closed, that all POSIX locks on the file, owned by the
process which closed the file descriptor, should be released.
The trick here is when those locks are released. The current code releases
all locks which exist when close is processing, but any locks in progress
are handled when the last reference to the open file is released.
There are three cases to consider.
One is the simple case, a multithreaded (mt) process has a file open and
races to close it and acquire a lock on it. In this case, the close will
release one reference to the open file and when the fcntl is done, it will
release the other reference. For this situation, no locks should exist on
the file when both the close and fcntl operations are done. The current
system will handle this case because the last reference to the open file is
being released.
The second case is when the mt process has dup(2)'d the file descriptor.
The close will release one reference to the file and the fcntl, when done,
will release another, but there will still be at least one more reference
to the open file. One could argue that the existence of a lock on the file
after the close has completed is okay, because it was acquired after the
close operation and there is still a way for the application to release the
lock on the file, using an existing file descriptor.
The third case is when the mt process has forked, after opening the file
and either before or after becoming an mt process. In this case, each
process would hold a reference to the open file. For each process, this
degenerates to first case above. However, the lock continues to exist
until both processes have released their references to the open file. This
lock could block other lock requests.
The changes to release the lock when the last reference to the open file
aren't quite right because they would allow the lock to exist as long as
there was a reference to the open file. This is too long.
The new proposed solution is to add support in the fcntl code path to
detect a race with close and then to release the lock which was just
acquired when such as race is detected. This causes locks to be released
in a timely fashion and for the system to conform to the POSIX semantic
specification.
This was tested by instrumenting a kernel to detect the handling locks and
then running a program which generates case #3 above. A dangling lock
could be reliably generated. When the changes to detect the close/fcntl
race were added, a dangling lock could no longer be generated.
Cc: Matthew Wilcox <willy@debian.org>
Cc: Trond Myklebust <trond.myklebust@fys.uio.no>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-07-27 12:45:09 -06:00
|
|
|
flock.l_type = F_UNLCK;
|
|
|
|
goto again;
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
|
[PATCH] stale POSIX lock handling
I believe that there is a problem with the handling of POSIX locks, which
the attached patch should address.
The problem appears to be a race between fcntl(2) and close(2). A
multithreaded application could close a file descriptor at the same time as
it is trying to acquire a lock using the same file descriptor. I would
suggest that that multithreaded application is not providing the proper
synchronization for itself, but the OS should still behave correctly.
SUS3 (Single UNIX Specification Version 3, read: POSIX) indicates that when
a file descriptor is closed, that all POSIX locks on the file, owned by the
process which closed the file descriptor, should be released.
The trick here is when those locks are released. The current code releases
all locks which exist when close is processing, but any locks in progress
are handled when the last reference to the open file is released.
There are three cases to consider.
One is the simple case, a multithreaded (mt) process has a file open and
races to close it and acquire a lock on it. In this case, the close will
release one reference to the open file and when the fcntl is done, it will
release the other reference. For this situation, no locks should exist on
the file when both the close and fcntl operations are done. The current
system will handle this case because the last reference to the open file is
being released.
The second case is when the mt process has dup(2)'d the file descriptor.
The close will release one reference to the file and the fcntl, when done,
will release another, but there will still be at least one more reference
to the open file. One could argue that the existence of a lock on the file
after the close has completed is okay, because it was acquired after the
close operation and there is still a way for the application to release the
lock on the file, using an existing file descriptor.
The third case is when the mt process has forked, after opening the file
and either before or after becoming an mt process. In this case, each
process would hold a reference to the open file. For each process, this
degenerates to first case above. However, the lock continues to exist
until both processes have released their references to the open file. This
lock could block other lock requests.
The changes to release the lock when the last reference to the open file
aren't quite right because they would allow the lock to exist as long as
there was a reference to the open file. This is too long.
The new proposed solution is to add support in the fcntl code path to
detect a race with close and then to release the lock which was just
acquired when such as race is detected. This causes locks to be released
in a timely fashion and for the system to conform to the POSIX semantic
specification.
This was tested by instrumenting a kernel to detect the handling locks and
then running a program which generates case #3 above. A dangling lock
could be reliably generated. When the changes to detect the close/fcntl
race were added, a dangling lock could no longer be generated.
Cc: Matthew Wilcox <willy@debian.org>
Cc: Trond Myklebust <trond.myklebust@fys.uio.no>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-07-27 12:45:09 -06:00
|
|
|
out:
|
2005-04-16 16:20:36 -06:00
|
|
|
locks_free_lock(file_lock);
|
|
|
|
return error;
|
|
|
|
}
|
|
|
|
|
|
|
|
#if BITS_PER_LONG == 32
|
|
|
|
/* Report the first existing lock that would conflict with l.
|
|
|
|
* This implements the F_GETLK command of fcntl().
|
|
|
|
*/
|
|
|
|
int fcntl_getlk64(struct file *filp, struct flock64 __user *l)
|
|
|
|
{
|
2007-02-20 22:55:18 -07:00
|
|
|
struct file_lock file_lock;
|
2005-04-16 16:20:36 -06:00
|
|
|
struct flock64 flock;
|
|
|
|
int error;
|
|
|
|
|
|
|
|
error = -EFAULT;
|
|
|
|
if (copy_from_user(&flock, l, sizeof(flock)))
|
|
|
|
goto out;
|
|
|
|
error = -EINVAL;
|
|
|
|
if ((flock.l_type != F_RDLCK) && (flock.l_type != F_WRLCK))
|
|
|
|
goto out;
|
|
|
|
|
|
|
|
error = flock64_to_posix_lock(filp, &file_lock, &flock);
|
|
|
|
if (error)
|
|
|
|
goto out;
|
|
|
|
|
2007-02-20 22:58:50 -07:00
|
|
|
error = vfs_test_lock(filp, &file_lock);
|
|
|
|
if (error)
|
|
|
|
goto out;
|
|
|
|
|
2007-02-20 22:55:18 -07:00
|
|
|
flock.l_type = file_lock.fl_type;
|
|
|
|
if (file_lock.fl_type != F_UNLCK)
|
|
|
|
posix_lock_to_flock64(&flock, &file_lock);
|
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
error = -EFAULT;
|
|
|
|
if (!copy_to_user(l, &flock, sizeof(flock)))
|
|
|
|
error = 0;
|
|
|
|
|
|
|
|
out:
|
|
|
|
return error;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Apply the lock described by l to an open file descriptor.
|
|
|
|
* This implements both the F_SETLK and F_SETLKW commands of fcntl().
|
|
|
|
*/
|
[PATCH] stale POSIX lock handling
I believe that there is a problem with the handling of POSIX locks, which
the attached patch should address.
The problem appears to be a race between fcntl(2) and close(2). A
multithreaded application could close a file descriptor at the same time as
it is trying to acquire a lock using the same file descriptor. I would
suggest that that multithreaded application is not providing the proper
synchronization for itself, but the OS should still behave correctly.
SUS3 (Single UNIX Specification Version 3, read: POSIX) indicates that when
a file descriptor is closed, that all POSIX locks on the file, owned by the
process which closed the file descriptor, should be released.
The trick here is when those locks are released. The current code releases
all locks which exist when close is processing, but any locks in progress
are handled when the last reference to the open file is released.
There are three cases to consider.
One is the simple case, a multithreaded (mt) process has a file open and
races to close it and acquire a lock on it. In this case, the close will
release one reference to the open file and when the fcntl is done, it will
release the other reference. For this situation, no locks should exist on
the file when both the close and fcntl operations are done. The current
system will handle this case because the last reference to the open file is
being released.
The second case is when the mt process has dup(2)'d the file descriptor.
The close will release one reference to the file and the fcntl, when done,
will release another, but there will still be at least one more reference
to the open file. One could argue that the existence of a lock on the file
after the close has completed is okay, because it was acquired after the
close operation and there is still a way for the application to release the
lock on the file, using an existing file descriptor.
The third case is when the mt process has forked, after opening the file
and either before or after becoming an mt process. In this case, each
process would hold a reference to the open file. For each process, this
degenerates to first case above. However, the lock continues to exist
until both processes have released their references to the open file. This
lock could block other lock requests.
The changes to release the lock when the last reference to the open file
aren't quite right because they would allow the lock to exist as long as
there was a reference to the open file. This is too long.
The new proposed solution is to add support in the fcntl code path to
detect a race with close and then to release the lock which was just
acquired when such as race is detected. This causes locks to be released
in a timely fashion and for the system to conform to the POSIX semantic
specification.
This was tested by instrumenting a kernel to detect the handling locks and
then running a program which generates case #3 above. A dangling lock
could be reliably generated. When the changes to detect the close/fcntl
race were added, a dangling lock could no longer be generated.
Cc: Matthew Wilcox <willy@debian.org>
Cc: Trond Myklebust <trond.myklebust@fys.uio.no>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-07-27 12:45:09 -06:00
|
|
|
int fcntl_setlk64(unsigned int fd, struct file *filp, unsigned int cmd,
|
|
|
|
struct flock64 __user *l)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
|
|
|
struct file_lock *file_lock = locks_alloc_lock();
|
|
|
|
struct flock64 flock;
|
|
|
|
struct inode *inode;
|
2008-05-06 11:58:34 -06:00
|
|
|
struct file *f;
|
2005-04-16 16:20:36 -06:00
|
|
|
int error;
|
|
|
|
|
|
|
|
if (file_lock == NULL)
|
|
|
|
return -ENOLCK;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* This might block, so we do it before checking the inode.
|
|
|
|
*/
|
|
|
|
error = -EFAULT;
|
|
|
|
if (copy_from_user(&flock, l, sizeof(flock)))
|
|
|
|
goto out;
|
|
|
|
|
2006-12-08 03:36:35 -07:00
|
|
|
inode = filp->f_path.dentry->d_inode;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
/* Don't allow mandatory locks on files that may be memory mapped
|
|
|
|
* and shared.
|
|
|
|
*/
|
2007-10-01 15:41:11 -06:00
|
|
|
if (mandatory_lock(inode) && mapping_writably_mapped(filp->f_mapping)) {
|
2005-04-16 16:20:36 -06:00
|
|
|
error = -EAGAIN;
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
[PATCH] stale POSIX lock handling
I believe that there is a problem with the handling of POSIX locks, which
the attached patch should address.
The problem appears to be a race between fcntl(2) and close(2). A
multithreaded application could close a file descriptor at the same time as
it is trying to acquire a lock using the same file descriptor. I would
suggest that that multithreaded application is not providing the proper
synchronization for itself, but the OS should still behave correctly.
SUS3 (Single UNIX Specification Version 3, read: POSIX) indicates that when
a file descriptor is closed, that all POSIX locks on the file, owned by the
process which closed the file descriptor, should be released.
The trick here is when those locks are released. The current code releases
all locks which exist when close is processing, but any locks in progress
are handled when the last reference to the open file is released.
There are three cases to consider.
One is the simple case, a multithreaded (mt) process has a file open and
races to close it and acquire a lock on it. In this case, the close will
release one reference to the open file and when the fcntl is done, it will
release the other reference. For this situation, no locks should exist on
the file when both the close and fcntl operations are done. The current
system will handle this case because the last reference to the open file is
being released.
The second case is when the mt process has dup(2)'d the file descriptor.
The close will release one reference to the file and the fcntl, when done,
will release another, but there will still be at least one more reference
to the open file. One could argue that the existence of a lock on the file
after the close has completed is okay, because it was acquired after the
close operation and there is still a way for the application to release the
lock on the file, using an existing file descriptor.
The third case is when the mt process has forked, after opening the file
and either before or after becoming an mt process. In this case, each
process would hold a reference to the open file. For each process, this
degenerates to first case above. However, the lock continues to exist
until both processes have released their references to the open file. This
lock could block other lock requests.
The changes to release the lock when the last reference to the open file
aren't quite right because they would allow the lock to exist as long as
there was a reference to the open file. This is too long.
The new proposed solution is to add support in the fcntl code path to
detect a race with close and then to release the lock which was just
acquired when such as race is detected. This causes locks to be released
in a timely fashion and for the system to conform to the POSIX semantic
specification.
This was tested by instrumenting a kernel to detect the handling locks and
then running a program which generates case #3 above. A dangling lock
could be reliably generated. When the changes to detect the close/fcntl
race were added, a dangling lock could no longer be generated.
Cc: Matthew Wilcox <willy@debian.org>
Cc: Trond Myklebust <trond.myklebust@fys.uio.no>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-07-27 12:45:09 -06:00
|
|
|
again:
|
2005-04-16 16:20:36 -06:00
|
|
|
error = flock64_to_posix_lock(filp, file_lock, &flock);
|
|
|
|
if (error)
|
|
|
|
goto out;
|
|
|
|
if (cmd == F_SETLKW64) {
|
|
|
|
file_lock->fl_flags |= FL_SLEEP;
|
|
|
|
}
|
|
|
|
|
|
|
|
error = -EBADF;
|
|
|
|
switch (flock.l_type) {
|
|
|
|
case F_RDLCK:
|
|
|
|
if (!(filp->f_mode & FMODE_READ))
|
|
|
|
goto out;
|
|
|
|
break;
|
|
|
|
case F_WRLCK:
|
|
|
|
if (!(filp->f_mode & FMODE_WRITE))
|
|
|
|
goto out;
|
|
|
|
break;
|
|
|
|
case F_UNLCK:
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
error = -EINVAL;
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
2008-07-25 02:48:57 -06:00
|
|
|
error = do_lock_file_wait(filp, cmd, file_lock);
|
2005-04-16 16:20:36 -06:00
|
|
|
|
[PATCH] stale POSIX lock handling
I believe that there is a problem with the handling of POSIX locks, which
the attached patch should address.
The problem appears to be a race between fcntl(2) and close(2). A
multithreaded application could close a file descriptor at the same time as
it is trying to acquire a lock using the same file descriptor. I would
suggest that that multithreaded application is not providing the proper
synchronization for itself, but the OS should still behave correctly.
SUS3 (Single UNIX Specification Version 3, read: POSIX) indicates that when
a file descriptor is closed, that all POSIX locks on the file, owned by the
process which closed the file descriptor, should be released.
The trick here is when those locks are released. The current code releases
all locks which exist when close is processing, but any locks in progress
are handled when the last reference to the open file is released.
There are three cases to consider.
One is the simple case, a multithreaded (mt) process has a file open and
races to close it and acquire a lock on it. In this case, the close will
release one reference to the open file and when the fcntl is done, it will
release the other reference. For this situation, no locks should exist on
the file when both the close and fcntl operations are done. The current
system will handle this case because the last reference to the open file is
being released.
The second case is when the mt process has dup(2)'d the file descriptor.
The close will release one reference to the file and the fcntl, when done,
will release another, but there will still be at least one more reference
to the open file. One could argue that the existence of a lock on the file
after the close has completed is okay, because it was acquired after the
close operation and there is still a way for the application to release the
lock on the file, using an existing file descriptor.
The third case is when the mt process has forked, after opening the file
and either before or after becoming an mt process. In this case, each
process would hold a reference to the open file. For each process, this
degenerates to first case above. However, the lock continues to exist
until both processes have released their references to the open file. This
lock could block other lock requests.
The changes to release the lock when the last reference to the open file
aren't quite right because they would allow the lock to exist as long as
there was a reference to the open file. This is too long.
The new proposed solution is to add support in the fcntl code path to
detect a race with close and then to release the lock which was just
acquired when such as race is detected. This causes locks to be released
in a timely fashion and for the system to conform to the POSIX semantic
specification.
This was tested by instrumenting a kernel to detect the handling locks and
then running a program which generates case #3 above. A dangling lock
could be reliably generated. When the changes to detect the close/fcntl
race were added, a dangling lock could no longer be generated.
Cc: Matthew Wilcox <willy@debian.org>
Cc: Trond Myklebust <trond.myklebust@fys.uio.no>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-07-27 12:45:09 -06:00
|
|
|
/*
|
|
|
|
* Attempt to detect a close/fcntl race and recover by
|
|
|
|
* releasing the lock that was just acquired.
|
|
|
|
*/
|
2008-05-06 11:58:34 -06:00
|
|
|
spin_lock(¤t->files->file_lock);
|
|
|
|
f = fcheck(fd);
|
|
|
|
spin_unlock(¤t->files->file_lock);
|
|
|
|
if (!error && f != filp && flock.l_type != F_UNLCK) {
|
[PATCH] stale POSIX lock handling
I believe that there is a problem with the handling of POSIX locks, which
the attached patch should address.
The problem appears to be a race between fcntl(2) and close(2). A
multithreaded application could close a file descriptor at the same time as
it is trying to acquire a lock using the same file descriptor. I would
suggest that that multithreaded application is not providing the proper
synchronization for itself, but the OS should still behave correctly.
SUS3 (Single UNIX Specification Version 3, read: POSIX) indicates that when
a file descriptor is closed, that all POSIX locks on the file, owned by the
process which closed the file descriptor, should be released.
The trick here is when those locks are released. The current code releases
all locks which exist when close is processing, but any locks in progress
are handled when the last reference to the open file is released.
There are three cases to consider.
One is the simple case, a multithreaded (mt) process has a file open and
races to close it and acquire a lock on it. In this case, the close will
release one reference to the open file and when the fcntl is done, it will
release the other reference. For this situation, no locks should exist on
the file when both the close and fcntl operations are done. The current
system will handle this case because the last reference to the open file is
being released.
The second case is when the mt process has dup(2)'d the file descriptor.
The close will release one reference to the file and the fcntl, when done,
will release another, but there will still be at least one more reference
to the open file. One could argue that the existence of a lock on the file
after the close has completed is okay, because it was acquired after the
close operation and there is still a way for the application to release the
lock on the file, using an existing file descriptor.
The third case is when the mt process has forked, after opening the file
and either before or after becoming an mt process. In this case, each
process would hold a reference to the open file. For each process, this
degenerates to first case above. However, the lock continues to exist
until both processes have released their references to the open file. This
lock could block other lock requests.
The changes to release the lock when the last reference to the open file
aren't quite right because they would allow the lock to exist as long as
there was a reference to the open file. This is too long.
The new proposed solution is to add support in the fcntl code path to
detect a race with close and then to release the lock which was just
acquired when such as race is detected. This causes locks to be released
in a timely fashion and for the system to conform to the POSIX semantic
specification.
This was tested by instrumenting a kernel to detect the handling locks and
then running a program which generates case #3 above. A dangling lock
could be reliably generated. When the changes to detect the close/fcntl
race were added, a dangling lock could no longer be generated.
Cc: Matthew Wilcox <willy@debian.org>
Cc: Trond Myklebust <trond.myklebust@fys.uio.no>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-07-27 12:45:09 -06:00
|
|
|
flock.l_type = F_UNLCK;
|
|
|
|
goto again;
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
|
|
|
|
out:
|
|
|
|
locks_free_lock(file_lock);
|
|
|
|
return error;
|
|
|
|
}
|
|
|
|
#endif /* BITS_PER_LONG == 32 */
|
|
|
|
|
|
|
|
/*
|
|
|
|
* This function is called when the file is being removed
|
|
|
|
* from the task's fd array. POSIX locks belonging to this task
|
|
|
|
* are deleted at this time.
|
|
|
|
*/
|
|
|
|
void locks_remove_posix(struct file *filp, fl_owner_t owner)
|
|
|
|
{
|
2006-06-23 03:05:11 -06:00
|
|
|
struct file_lock lock;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
/*
|
|
|
|
* If there are no locks held on this file, we don't need to call
|
|
|
|
* posix_lock_file(). Another process could be setting a lock on this
|
|
|
|
* file at the same time, but we wouldn't remove that lock anyway.
|
|
|
|
*/
|
2006-12-08 03:36:35 -07:00
|
|
|
if (!filp->f_path.dentry->d_inode->i_flock)
|
2005-04-16 16:20:36 -06:00
|
|
|
return;
|
|
|
|
|
|
|
|
lock.fl_type = F_UNLCK;
|
2006-06-23 03:05:12 -06:00
|
|
|
lock.fl_flags = FL_POSIX | FL_CLOSE;
|
2005-04-16 16:20:36 -06:00
|
|
|
lock.fl_start = 0;
|
|
|
|
lock.fl_end = OFFSET_MAX;
|
|
|
|
lock.fl_owner = owner;
|
|
|
|
lock.fl_pid = current->tgid;
|
|
|
|
lock.fl_file = filp;
|
|
|
|
lock.fl_ops = NULL;
|
|
|
|
lock.fl_lmops = NULL;
|
|
|
|
|
2007-01-18 14:15:35 -07:00
|
|
|
vfs_lock_file(filp, F_SETLK, &lock, NULL);
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
if (lock.fl_ops && lock.fl_ops->fl_release_private)
|
|
|
|
lock.fl_ops->fl_release_private(&lock);
|
|
|
|
}
|
|
|
|
|
|
|
|
EXPORT_SYMBOL(locks_remove_posix);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* This function is called on the last close of an open file.
|
|
|
|
*/
|
|
|
|
void locks_remove_flock(struct file *filp)
|
|
|
|
{
|
2006-12-08 03:36:35 -07:00
|
|
|
struct inode * inode = filp->f_path.dentry->d_inode;
|
2005-04-16 16:20:36 -06:00
|
|
|
struct file_lock *fl;
|
|
|
|
struct file_lock **before;
|
|
|
|
|
|
|
|
if (!inode->i_flock)
|
|
|
|
return;
|
|
|
|
|
|
|
|
if (filp->f_op && filp->f_op->flock) {
|
|
|
|
struct file_lock fl = {
|
|
|
|
.fl_pid = current->tgid,
|
|
|
|
.fl_file = filp,
|
|
|
|
.fl_flags = FL_FLOCK,
|
|
|
|
.fl_type = F_UNLCK,
|
|
|
|
.fl_end = OFFSET_MAX,
|
|
|
|
};
|
|
|
|
filp->f_op->flock(filp, F_SETLKW, &fl);
|
2005-06-22 11:16:31 -06:00
|
|
|
if (fl.fl_ops && fl.fl_ops->fl_release_private)
|
|
|
|
fl.fl_ops->fl_release_private(&fl);
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
|
2010-09-18 07:09:31 -06:00
|
|
|
lock_flocks();
|
2005-04-16 16:20:36 -06:00
|
|
|
before = &inode->i_flock;
|
|
|
|
|
|
|
|
while ((fl = *before) != NULL) {
|
|
|
|
if (fl->fl_file == filp) {
|
[PATCH] stale POSIX lock handling
I believe that there is a problem with the handling of POSIX locks, which
the attached patch should address.
The problem appears to be a race between fcntl(2) and close(2). A
multithreaded application could close a file descriptor at the same time as
it is trying to acquire a lock using the same file descriptor. I would
suggest that that multithreaded application is not providing the proper
synchronization for itself, but the OS should still behave correctly.
SUS3 (Single UNIX Specification Version 3, read: POSIX) indicates that when
a file descriptor is closed, that all POSIX locks on the file, owned by the
process which closed the file descriptor, should be released.
The trick here is when those locks are released. The current code releases
all locks which exist when close is processing, but any locks in progress
are handled when the last reference to the open file is released.
There are three cases to consider.
One is the simple case, a multithreaded (mt) process has a file open and
races to close it and acquire a lock on it. In this case, the close will
release one reference to the open file and when the fcntl is done, it will
release the other reference. For this situation, no locks should exist on
the file when both the close and fcntl operations are done. The current
system will handle this case because the last reference to the open file is
being released.
The second case is when the mt process has dup(2)'d the file descriptor.
The close will release one reference to the file and the fcntl, when done,
will release another, but there will still be at least one more reference
to the open file. One could argue that the existence of a lock on the file
after the close has completed is okay, because it was acquired after the
close operation and there is still a way for the application to release the
lock on the file, using an existing file descriptor.
The third case is when the mt process has forked, after opening the file
and either before or after becoming an mt process. In this case, each
process would hold a reference to the open file. For each process, this
degenerates to first case above. However, the lock continues to exist
until both processes have released their references to the open file. This
lock could block other lock requests.
The changes to release the lock when the last reference to the open file
aren't quite right because they would allow the lock to exist as long as
there was a reference to the open file. This is too long.
The new proposed solution is to add support in the fcntl code path to
detect a race with close and then to release the lock which was just
acquired when such as race is detected. This causes locks to be released
in a timely fashion and for the system to conform to the POSIX semantic
specification.
This was tested by instrumenting a kernel to detect the handling locks and
then running a program which generates case #3 above. A dangling lock
could be reliably generated. When the changes to detect the close/fcntl
race were added, a dangling lock could no longer be generated.
Cc: Matthew Wilcox <willy@debian.org>
Cc: Trond Myklebust <trond.myklebust@fys.uio.no>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-07-27 12:45:09 -06:00
|
|
|
if (IS_FLOCK(fl)) {
|
2005-04-16 16:20:36 -06:00
|
|
|
locks_delete_lock(before);
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
if (IS_LEASE(fl)) {
|
|
|
|
lease_modify(before, F_UNLCK);
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
/* What? */
|
|
|
|
BUG();
|
|
|
|
}
|
|
|
|
before = &fl->fl_next;
|
|
|
|
}
|
2010-09-18 07:09:31 -06:00
|
|
|
unlock_flocks();
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* posix_unblock_lock - stop waiting for a file lock
|
|
|
|
* @filp: how the file was opened
|
|
|
|
* @waiter: the lock which was waiting
|
|
|
|
*
|
|
|
|
* lockd needs to block waiting for locks.
|
|
|
|
*/
|
2006-01-03 01:55:46 -07:00
|
|
|
int
|
2005-04-16 16:20:36 -06:00
|
|
|
posix_unblock_lock(struct file *filp, struct file_lock *waiter)
|
|
|
|
{
|
2006-01-03 01:55:46 -07:00
|
|
|
int status = 0;
|
|
|
|
|
2010-09-18 07:09:31 -06:00
|
|
|
lock_flocks();
|
2006-01-03 01:55:44 -07:00
|
|
|
if (waiter->fl_next)
|
2005-04-16 16:20:36 -06:00
|
|
|
__locks_delete_block(waiter);
|
2006-01-03 01:55:46 -07:00
|
|
|
else
|
|
|
|
status = -ENOENT;
|
2010-09-18 07:09:31 -06:00
|
|
|
unlock_flocks();
|
2006-01-03 01:55:46 -07:00
|
|
|
return status;
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
|
|
|
|
EXPORT_SYMBOL(posix_unblock_lock);
|
|
|
|
|
2007-01-18 15:52:58 -07:00
|
|
|
/**
|
|
|
|
* vfs_cancel_lock - file byte range unblock lock
|
|
|
|
* @filp: The file to apply the unblock to
|
|
|
|
* @fl: The lock to be unblocked
|
|
|
|
*
|
|
|
|
* Used by lock managers to cancel blocked requests
|
|
|
|
*/
|
|
|
|
int vfs_cancel_lock(struct file *filp, struct file_lock *fl)
|
|
|
|
{
|
|
|
|
if (filp->f_op && filp->f_op->lock)
|
|
|
|
return filp->f_op->lock(filp, F_CANCELLK, fl);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
EXPORT_SYMBOL_GPL(vfs_cancel_lock);
|
|
|
|
|
2007-10-01 15:41:15 -06:00
|
|
|
#ifdef CONFIG_PROC_FS
|
2008-10-04 12:34:18 -06:00
|
|
|
#include <linux/proc_fs.h>
|
2007-10-01 15:41:15 -06:00
|
|
|
#include <linux/seq_file.h>
|
|
|
|
|
|
|
|
static void lock_get_status(struct seq_file *f, struct file_lock *fl,
|
2010-10-26 15:22:33 -06:00
|
|
|
loff_t id, char *pfx)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
|
|
|
struct inode *inode = NULL;
|
2008-01-16 17:07:08 -07:00
|
|
|
unsigned int fl_pid;
|
|
|
|
|
|
|
|
if (fl->fl_nspid)
|
2008-02-08 05:19:20 -07:00
|
|
|
fl_pid = pid_vnr(fl->fl_nspid);
|
2008-01-16 17:07:08 -07:00
|
|
|
else
|
|
|
|
fl_pid = fl->fl_pid;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
if (fl->fl_file != NULL)
|
2006-12-08 03:36:35 -07:00
|
|
|
inode = fl->fl_file->f_path.dentry->d_inode;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2010-10-26 15:22:33 -06:00
|
|
|
seq_printf(f, "%lld:%s ", id, pfx);
|
2005-04-16 16:20:36 -06:00
|
|
|
if (IS_POSIX(fl)) {
|
2007-10-01 15:41:15 -06:00
|
|
|
seq_printf(f, "%6s %s ",
|
2005-04-16 16:20:36 -06:00
|
|
|
(fl->fl_flags & FL_ACCESS) ? "ACCESS" : "POSIX ",
|
|
|
|
(inode == NULL) ? "*NOINODE*" :
|
2007-10-01 15:41:11 -06:00
|
|
|
mandatory_lock(inode) ? "MANDATORY" : "ADVISORY ");
|
2005-04-16 16:20:36 -06:00
|
|
|
} else if (IS_FLOCK(fl)) {
|
|
|
|
if (fl->fl_type & LOCK_MAND) {
|
2007-10-01 15:41:15 -06:00
|
|
|
seq_printf(f, "FLOCK MSNFS ");
|
2005-04-16 16:20:36 -06:00
|
|
|
} else {
|
2007-10-01 15:41:15 -06:00
|
|
|
seq_printf(f, "FLOCK ADVISORY ");
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
} else if (IS_LEASE(fl)) {
|
2007-10-01 15:41:15 -06:00
|
|
|
seq_printf(f, "LEASE ");
|
2011-07-26 18:10:51 -06:00
|
|
|
if (lease_breaking(fl))
|
2007-10-01 15:41:15 -06:00
|
|
|
seq_printf(f, "BREAKING ");
|
2005-04-16 16:20:36 -06:00
|
|
|
else if (fl->fl_file)
|
2007-10-01 15:41:15 -06:00
|
|
|
seq_printf(f, "ACTIVE ");
|
2005-04-16 16:20:36 -06:00
|
|
|
else
|
2007-10-01 15:41:15 -06:00
|
|
|
seq_printf(f, "BREAKER ");
|
2005-04-16 16:20:36 -06:00
|
|
|
} else {
|
2007-10-01 15:41:15 -06:00
|
|
|
seq_printf(f, "UNKNOWN UNKNOWN ");
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
if (fl->fl_type & LOCK_MAND) {
|
2007-10-01 15:41:15 -06:00
|
|
|
seq_printf(f, "%s ",
|
2005-04-16 16:20:36 -06:00
|
|
|
(fl->fl_type & LOCK_READ)
|
|
|
|
? (fl->fl_type & LOCK_WRITE) ? "RW " : "READ "
|
|
|
|
: (fl->fl_type & LOCK_WRITE) ? "WRITE" : "NONE ");
|
|
|
|
} else {
|
2007-10-01 15:41:15 -06:00
|
|
|
seq_printf(f, "%s ",
|
2011-07-26 18:10:51 -06:00
|
|
|
(lease_breaking(fl))
|
2012-08-02 13:46:30 -06:00
|
|
|
? (fl->fl_type == F_UNLCK) ? "UNLCK" : "READ "
|
|
|
|
: (fl->fl_type == F_WRLCK) ? "WRITE" : "READ ");
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
if (inode) {
|
|
|
|
#ifdef WE_CAN_BREAK_LSLK_NOW
|
2008-01-16 17:07:08 -07:00
|
|
|
seq_printf(f, "%d %s:%ld ", fl_pid,
|
2005-04-16 16:20:36 -06:00
|
|
|
inode->i_sb->s_id, inode->i_ino);
|
|
|
|
#else
|
|
|
|
/* userspace relies on this representation of dev_t ;-( */
|
2008-01-16 17:07:08 -07:00
|
|
|
seq_printf(f, "%d %02x:%02x:%ld ", fl_pid,
|
2005-04-16 16:20:36 -06:00
|
|
|
MAJOR(inode->i_sb->s_dev),
|
|
|
|
MINOR(inode->i_sb->s_dev), inode->i_ino);
|
|
|
|
#endif
|
|
|
|
} else {
|
2008-01-16 17:07:08 -07:00
|
|
|
seq_printf(f, "%d <none>:0 ", fl_pid);
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
if (IS_POSIX(fl)) {
|
|
|
|
if (fl->fl_end == OFFSET_MAX)
|
2007-10-01 15:41:15 -06:00
|
|
|
seq_printf(f, "%Ld EOF\n", fl->fl_start);
|
2005-04-16 16:20:36 -06:00
|
|
|
else
|
2007-10-01 15:41:15 -06:00
|
|
|
seq_printf(f, "%Ld %Ld\n", fl->fl_start, fl->fl_end);
|
2005-04-16 16:20:36 -06:00
|
|
|
} else {
|
2007-10-01 15:41:15 -06:00
|
|
|
seq_printf(f, "0 EOF\n");
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2007-10-01 15:41:15 -06:00
|
|
|
static int locks_show(struct seq_file *f, void *v)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
2007-10-01 15:41:15 -06:00
|
|
|
struct file_lock *fl, *bfl;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2007-10-01 15:41:15 -06:00
|
|
|
fl = list_entry(v, struct file_lock, fl_link);
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2010-10-26 15:22:33 -06:00
|
|
|
lock_get_status(f, fl, *((loff_t *)f->private), "");
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2007-10-01 15:41:15 -06:00
|
|
|
list_for_each_entry(bfl, &fl->fl_block, fl_block)
|
2010-10-26 15:22:33 -06:00
|
|
|
lock_get_status(f, bfl, *((loff_t *)f->private), " ->");
|
2007-10-02 12:21:34 -06:00
|
|
|
|
2007-10-01 15:41:15 -06:00
|
|
|
return 0;
|
|
|
|
}
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2007-10-01 15:41:15 -06:00
|
|
|
static void *locks_start(struct seq_file *f, loff_t *pos)
|
|
|
|
{
|
2010-10-26 15:22:33 -06:00
|
|
|
loff_t *p = f->private;
|
|
|
|
|
2010-09-18 07:09:31 -06:00
|
|
|
lock_flocks();
|
2010-10-26 15:22:33 -06:00
|
|
|
*p = (*pos + 1);
|
2007-10-01 15:41:15 -06:00
|
|
|
return seq_list_start(&file_lock_list, *pos);
|
|
|
|
}
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2007-10-01 15:41:15 -06:00
|
|
|
static void *locks_next(struct seq_file *f, void *v, loff_t *pos)
|
|
|
|
{
|
2010-10-26 15:22:33 -06:00
|
|
|
loff_t *p = f->private;
|
|
|
|
++*p;
|
2007-10-01 15:41:15 -06:00
|
|
|
return seq_list_next(v, &file_lock_list, pos);
|
|
|
|
}
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2007-10-01 15:41:15 -06:00
|
|
|
static void locks_stop(struct seq_file *f, void *v)
|
|
|
|
{
|
2010-09-18 07:09:31 -06:00
|
|
|
unlock_flocks();
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
|
2008-10-04 12:34:18 -06:00
|
|
|
static const struct seq_operations locks_seq_operations = {
|
2007-10-01 15:41:15 -06:00
|
|
|
.start = locks_start,
|
|
|
|
.next = locks_next,
|
|
|
|
.stop = locks_stop,
|
|
|
|
.show = locks_show,
|
|
|
|
};
|
2008-10-04 12:34:18 -06:00
|
|
|
|
|
|
|
static int locks_open(struct inode *inode, struct file *filp)
|
|
|
|
{
|
2010-10-26 15:22:33 -06:00
|
|
|
return seq_open_private(filp, &locks_seq_operations, sizeof(loff_t));
|
2008-10-04 12:34:18 -06:00
|
|
|
}
|
|
|
|
|
|
|
|
static const struct file_operations proc_locks_operations = {
|
|
|
|
.open = locks_open,
|
|
|
|
.read = seq_read,
|
|
|
|
.llseek = seq_lseek,
|
2010-10-26 15:22:33 -06:00
|
|
|
.release = seq_release_private,
|
2008-10-04 12:34:18 -06:00
|
|
|
};
|
|
|
|
|
|
|
|
static int __init proc_locks_init(void)
|
|
|
|
{
|
|
|
|
proc_create("locks", 0, NULL, &proc_locks_operations);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
module_init(proc_locks_init);
|
2007-10-01 15:41:15 -06:00
|
|
|
#endif
|
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
/**
|
|
|
|
* lock_may_read - checks that the region is free of locks
|
|
|
|
* @inode: the inode that is being read
|
|
|
|
* @start: the first byte to read
|
|
|
|
* @len: the number of bytes to read
|
|
|
|
*
|
|
|
|
* Emulates Windows locking requirements. Whole-file
|
|
|
|
* mandatory locks (share modes) can prohibit a read and
|
|
|
|
* byte-range POSIX locks can prohibit a read if they overlap.
|
|
|
|
*
|
|
|
|
* N.B. this function is only ever called
|
|
|
|
* from knfsd and ownership of locks is never checked.
|
|
|
|
*/
|
|
|
|
int lock_may_read(struct inode *inode, loff_t start, unsigned long len)
|
|
|
|
{
|
|
|
|
struct file_lock *fl;
|
|
|
|
int result = 1;
|
2010-09-18 07:09:31 -06:00
|
|
|
lock_flocks();
|
2005-04-16 16:20:36 -06:00
|
|
|
for (fl = inode->i_flock; fl != NULL; fl = fl->fl_next) {
|
|
|
|
if (IS_POSIX(fl)) {
|
|
|
|
if (fl->fl_type == F_RDLCK)
|
|
|
|
continue;
|
|
|
|
if ((fl->fl_end < start) || (fl->fl_start > (start + len)))
|
|
|
|
continue;
|
|
|
|
} else if (IS_FLOCK(fl)) {
|
|
|
|
if (!(fl->fl_type & LOCK_MAND))
|
|
|
|
continue;
|
|
|
|
if (fl->fl_type & LOCK_READ)
|
|
|
|
continue;
|
|
|
|
} else
|
|
|
|
continue;
|
|
|
|
result = 0;
|
|
|
|
break;
|
|
|
|
}
|
2010-09-18 07:09:31 -06:00
|
|
|
unlock_flocks();
|
2005-04-16 16:20:36 -06:00
|
|
|
return result;
|
|
|
|
}
|
|
|
|
|
|
|
|
EXPORT_SYMBOL(lock_may_read);
|
|
|
|
|
|
|
|
/**
|
|
|
|
* lock_may_write - checks that the region is free of locks
|
|
|
|
* @inode: the inode that is being written
|
|
|
|
* @start: the first byte to write
|
|
|
|
* @len: the number of bytes to write
|
|
|
|
*
|
|
|
|
* Emulates Windows locking requirements. Whole-file
|
|
|
|
* mandatory locks (share modes) can prohibit a write and
|
|
|
|
* byte-range POSIX locks can prohibit a write if they overlap.
|
|
|
|
*
|
|
|
|
* N.B. this function is only ever called
|
|
|
|
* from knfsd and ownership of locks is never checked.
|
|
|
|
*/
|
|
|
|
int lock_may_write(struct inode *inode, loff_t start, unsigned long len)
|
|
|
|
{
|
|
|
|
struct file_lock *fl;
|
|
|
|
int result = 1;
|
2010-09-18 07:09:31 -06:00
|
|
|
lock_flocks();
|
2005-04-16 16:20:36 -06:00
|
|
|
for (fl = inode->i_flock; fl != NULL; fl = fl->fl_next) {
|
|
|
|
if (IS_POSIX(fl)) {
|
|
|
|
if ((fl->fl_end < start) || (fl->fl_start > (start + len)))
|
|
|
|
continue;
|
|
|
|
} else if (IS_FLOCK(fl)) {
|
|
|
|
if (!(fl->fl_type & LOCK_MAND))
|
|
|
|
continue;
|
|
|
|
if (fl->fl_type & LOCK_WRITE)
|
|
|
|
continue;
|
|
|
|
} else
|
|
|
|
continue;
|
|
|
|
result = 0;
|
|
|
|
break;
|
|
|
|
}
|
2010-09-18 07:09:31 -06:00
|
|
|
unlock_flocks();
|
2005-04-16 16:20:36 -06:00
|
|
|
return result;
|
|
|
|
}
|
|
|
|
|
|
|
|
EXPORT_SYMBOL(lock_may_write);
|
|
|
|
|
|
|
|
static int __init filelock_init(void)
|
|
|
|
{
|
|
|
|
filelock_cache = kmem_cache_create("file_lock_cache",
|
2011-07-07 05:06:09 -06:00
|
|
|
sizeof(struct file_lock), 0, SLAB_PANIC, NULL);
|
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
core_initcall(filelock_init);
|