svclock.c

Linux中关于远程文件锁定的支持的源代码

/*
 * Test for presence of a conflicting lock.
 */
__be32
nlmsvc_testlock(struct svc_rqst *rqstp, struct nlm_file *file,
		struct nlm_host *host, struct nlm_lock *lock,
		struct nlm_lock *conflock, struct nlm_cookie *cookie)
{
	struct nlm_block 	*block = NULL;
	int			error;
	__be32			ret;

	dprintk("lockd: nlmsvc_testlock(%s/%ld, ty=%d, %Ld-%Ld)\n",
				file->f_file->f_path.dentry->d_inode->i_sb->s_id,
				file->f_file->f_path.dentry->d_inode->i_ino,
				lock->fl.fl_type,
				(long long)lock->fl.fl_start,
				(long long)lock->fl.fl_end);

	/* Get existing block (in case client is busy-waiting) */
	block = nlmsvc_lookup_block(file, lock);

	if (block == NULL) {
		struct file_lock *conf = kzalloc(sizeof(*conf), GFP_KERNEL);

		if (conf == NULL)
			return nlm_granted;
		block = nlmsvc_create_block(rqstp, host, file, lock, cookie);
		if (block == NULL) {
			kfree(conf);
			return nlm_granted;
		}
		block->b_fl = conf;
	}
	if (block->b_flags & B_QUEUED) {
		dprintk("lockd: nlmsvc_testlock deferred block %p flags %d fl %p\n",
			block, block->b_flags, block->b_fl);
		if (block->b_flags & B_TIMED_OUT) {
			nlmsvc_unlink_block(block);
			ret = nlm_lck_denied;
			goto out;
		}
		if (block->b_flags & B_GOT_CALLBACK) {
			nlmsvc_unlink_block(block);
			if (block->b_fl != NULL
					&& block->b_fl->fl_type != F_UNLCK) {
				lock->fl = *block->b_fl;
				goto conf_lock;
			} else {
				ret = nlm_granted;
				goto out;
			}
		}
		ret = nlm_drop_reply;
		goto out;
	}

	if (locks_in_grace()) {
		ret = nlm_lck_denied_grace_period;
		goto out;
	}
	error = vfs_test_lock(file->f_file, &lock->fl);
	if (error == FILE_LOCK_DEFERRED) {
		ret = nlmsvc_defer_lock_rqst(rqstp, block);
		goto out;
	}
	if (error) {
		ret = nlm_lck_denied_nolocks;
		goto out;
	}
	if (lock->fl.fl_type == F_UNLCK) {
		ret = nlm_granted;
		goto out;
	}

conf_lock:
	dprintk("lockd: conflicting lock(ty=%d, %Ld-%Ld)\n",
		lock->fl.fl_type, (long long)lock->fl.fl_start,
		(long long)lock->fl.fl_end);
	conflock->caller = "somehost";	/* FIXME */
	conflock->len = strlen(conflock->caller);
	conflock->oh.len = 0;		/* don't return OH info */
	conflock->svid = lock->fl.fl_pid;
	conflock->fl.fl_type = lock->fl.fl_type;
	conflock->fl.fl_start = lock->fl.fl_start;
	conflock->fl.fl_end = lock->fl.fl_end;
	ret = nlm_lck_denied;
out:
	if (block)
		nlmsvc_release_block(block);
	return ret;
}

/*
 * Remove a lock.
 * This implies a CANCEL call: We send a GRANT_MSG, the client replies
 * with a GRANT_RES call which gets lost, and calls UNLOCK immediately
 * afterwards. In this case the block will still be there, and hence
 * must be removed.
 */
__be32
nlmsvc_unlock(struct nlm_file *file, struct nlm_lock *lock)
{
	int	error;

	dprintk("lockd: nlmsvc_unlock(%s/%ld, pi=%d, %Ld-%Ld)\n",
				file->f_file->f_path.dentry->d_inode->i_sb->s_id,
				file->f_file->f_path.dentry->d_inode->i_ino,
				lock->fl.fl_pid,
				(long long)lock->fl.fl_start,
				(long long)lock->fl.fl_end);

	/* First, cancel any lock that might be there */
	nlmsvc_cancel_blocked(file, lock);

	lock->fl.fl_type = F_UNLCK;
	error = vfs_lock_file(file->f_file, F_SETLK, &lock->fl, NULL);

	return (error < 0)? nlm_lck_denied_nolocks : nlm_granted;
}

/*
 * Cancel a previously blocked request.
 *
 * A cancel request always overrides any grant that may currently
 * be in progress.
 * The calling procedure must check whether the file can be closed.
 */
__be32
nlmsvc_cancel_blocked(struct nlm_file *file, struct nlm_lock *lock)
{
	struct nlm_block	*block;
	int status = 0;

	dprintk("lockd: nlmsvc_cancel(%s/%ld, pi=%d, %Ld-%Ld)\n",
				file->f_file->f_path.dentry->d_inode->i_sb->s_id,
				file->f_file->f_path.dentry->d_inode->i_ino,
				lock->fl.fl_pid,
				(long long)lock->fl.fl_start,
				(long long)lock->fl.fl_end);

	if (locks_in_grace())
		return nlm_lck_denied_grace_period;

	mutex_lock(&file->f_mutex);
	block = nlmsvc_lookup_block(file, lock);
	mutex_unlock(&file->f_mutex);
	if (block != NULL) {
		vfs_cancel_lock(block->b_file->f_file,
				&block->b_call->a_args.lock.fl);
		status = nlmsvc_unlink_block(block);
		nlmsvc_release_block(block);
	}
	return status ? nlm_lck_denied : nlm_granted;
}

/*
 * This is a callback from the filesystem for VFS file lock requests.
 * It will be used if fl_grant is defined and the filesystem can not
 * respond to the request immediately.
 * For GETLK request it will copy the reply to the nlm_block.
 * For SETLK or SETLKW request it will get the local posix lock.
 * In all cases it will move the block to the head of nlm_blocked q where
 * nlmsvc_retry_blocked() can send back a reply for SETLKW or revisit the
 * deferred rpc for GETLK and SETLK.
 */
static void
nlmsvc_update_deferred_block(struct nlm_block *block, struct file_lock *conf,
			     int result)
{
	block->b_flags |= B_GOT_CALLBACK;
	if (result == 0)
		block->b_granted = 1;
	else
		block->b_flags |= B_TIMED_OUT;
	if (conf) {
		if (block->b_fl)
			__locks_copy_lock(block->b_fl, conf);
	}
}

static int nlmsvc_grant_deferred(struct file_lock *fl, struct file_lock *conf,
					int result)
{
	struct nlm_block *block;
	int rc = -ENOENT;

	lock_kernel();
	list_for_each_entry(block, &nlm_blocked, b_list) {
		if (nlm_compare_locks(&block->b_call->a_args.lock.fl, fl)) {
			dprintk("lockd: nlmsvc_notify_blocked block %p flags %d\n",
							block, block->b_flags);
			if (block->b_flags & B_QUEUED) {
				if (block->b_flags & B_TIMED_OUT) {
					rc = -ENOLCK;
					break;
				}
				nlmsvc_update_deferred_block(block, conf, result);
			} else if (result == 0)
				block->b_granted = 1;

			nlmsvc_insert_block(block, 0);
			svc_wake_up(block->b_daemon);
			rc = 0;
			break;
		}
	}
	unlock_kernel();
	if (rc == -ENOENT)
		printk(KERN_WARNING "lockd: grant for unknown block\n");
	return rc;
}

/*
 * Unblock a blocked lock request. This is a callback invoked from the
 * VFS layer when a lock on which we blocked is removed.
 *
 * This function doesn't grant the blocked lock instantly, but rather moves
 * the block to the head of nlm_blocked where it can be picked up by lockd.
 */
static void
nlmsvc_notify_blocked(struct file_lock *fl)
{
	struct nlm_block	*block;

	dprintk("lockd: VFS unblock notification for block %p\n", fl);
	list_for_each_entry(block, &nlm_blocked, b_list) {
		if (nlm_compare_locks(&block->b_call->a_args.lock.fl, fl)) {
			nlmsvc_insert_block(block, 0);
			svc_wake_up(block->b_daemon);
			return;
		}
	}

	printk(KERN_WARNING "lockd: notification for unknown block!\n");
}

static int nlmsvc_same_owner(struct file_lock *fl1, struct file_lock *fl2)
{
	return fl1->fl_owner == fl2->fl_owner && fl1->fl_pid == fl2->fl_pid;
}

struct lock_manager_operations nlmsvc_lock_operations = {
	.fl_compare_owner = nlmsvc_same_owner,
	.fl_notify = nlmsvc_notify_blocked,
	.fl_grant = nlmsvc_grant_deferred,
};

/*
 * Try to claim a lock that was previously blocked.
 *
 * Note that we use both the RPC_GRANTED_MSG call _and_ an async
 * RPC thread when notifying the client. This seems like overkill...
 * Here's why:
 *  -	we don't want to use a synchronous RPC thread, otherwise
 *	we might find ourselves hanging on a dead portmapper.
 *  -	Some lockd implementations (e.g. HP) don't react to
 *	RPC_GRANTED calls; they seem to insist on RPC_GRANTED_MSG calls.
 */
static void
nlmsvc_grant_blocked(struct nlm_block *block)
{
	struct nlm_file		*file = block->b_file;
	struct nlm_lock		*lock = &block->b_call->a_args.lock;
	int			error;

	dprintk("lockd: grant blocked lock %p\n", block);

	kref_get(&block->b_count);

	/* Unlink block request from list */
	nlmsvc_unlink_block(block);

	/* If b_granted is true this means we've been here before.
	 * Just retry the grant callback, possibly refreshing the RPC
	 * binding */
	if (block->b_granted) {
		nlm_rebind_host(block->b_host);
		goto callback;
	}

	/* Try the lock operation again */
	lock->fl.fl_flags |= FL_SLEEP;
	error = vfs_lock_file(file->f_file, F_SETLK, &lock->fl, NULL);
	lock->fl.fl_flags &= ~FL_SLEEP;

	switch (error) {
	case 0:
		break;
	case FILE_LOCK_DEFERRED:
		dprintk("lockd: lock still blocked error %d\n", error);
		nlmsvc_insert_block(block, NLM_NEVER);
		nlmsvc_release_block(block);
		return;
	default:
		printk(KERN_WARNING "lockd: unexpected error %d in %s!\n",
				-error, __func__);
		nlmsvc_insert_block(block, 10 * HZ);
		nlmsvc_release_block(block);
		return;
	}

callback:
	/* Lock was granted by VFS. */
	dprintk("lockd: GRANTing blocked lock.\n");
	block->b_granted = 1;

	/* keep block on the list, but don't reattempt until the RPC
	 * completes or the submission fails
	 */
	nlmsvc_insert_block(block, NLM_NEVER);

	/* Call the client -- use a soft RPC task since nlmsvc_retry_blocked
	 * will queue up a new one if this one times out
	 */
	error = nlm_async_call(block->b_call, NLMPROC_GRANTED_MSG,
				&nlmsvc_grant_ops);

	/* RPC submission failed, wait a bit and retry */
	if (error < 0)
		nlmsvc_insert_block(block, 10 * HZ);
}

/*
 * This is the callback from the RPC layer when the NLM_GRANTED_MSG
 * RPC call has succeeded or timed out.
 * Like all RPC callbacks, it is invoked by the rpciod process, so it
 * better not sleep. Therefore, we put the blocked lock on the nlm_blocked
 * chain once more in order to have it removed by lockd itself (which can
 * then sleep on the file semaphore without disrupting e.g. the nfs client).
 */
static void nlmsvc_grant_callback(struct rpc_task *task, void *data)
{
	struct nlm_rqst		*call = data;
	struct nlm_block	*block = call->a_block;
	unsigned long		timeout;

	dprintk("lockd: GRANT_MSG RPC callback\n");

	lock_kernel();
	/* if the block is not on a list at this point then it has
	 * been invalidated. Don't try to requeue it.
	 *
	 * FIXME: it's possible that the block is removed from the list
	 * after this check but before the nlmsvc_insert_block. In that
	 * case it will be added back. Perhaps we need better locking
	 * for nlm_blocked?
	 */
	if (list_empty(&block->b_list))
		goto out;

	/* Technically, we should down the file semaphore here. Since we
	 * move the block towards the head of the queue only, no harm
	 * can be done, though. */
	if (task->tk_status < 0) {
		/* RPC error: Re-insert for retransmission */
		timeout = 10 * HZ;
	} else {
		/* Call was successful, now wait for client callback */
		timeout = 60 * HZ;
	}
	nlmsvc_insert_block(block, timeout);
	svc_wake_up(block->b_daemon);
out:
	unlock_kernel();
}

static void nlmsvc_grant_release(void *data)
{
	struct nlm_rqst		*call = data;

	lock_kernel();
	nlmsvc_release_block(call->a_block);
	unlock_kernel();
}

static const struct rpc_call_ops nlmsvc_grant_ops = {
	.rpc_call_done = nlmsvc_grant_callback,
	.rpc_release = nlmsvc_grant_release,
};

/*
 * We received a GRANT_RES callback. Try to find the corresponding
 * block.
 */
void
nlmsvc_grant_reply(struct nlm_cookie *cookie, __be32 status)
{
	struct nlm_block	*block;

	dprintk("grant_reply: looking for cookie %x, s=%d \n",
		*(unsigned int *)(cookie->data), status);
	if (!(block = nlmsvc_find_block(cookie)))
		return;

	if (block) {
		if (status == nlm_lck_denied_grace_period) {
			/* Try again in a couple of seconds */
			nlmsvc_insert_block(block, 10 * HZ);
		} else {
			/* Lock is now held by client, or has been rejected.
			 * In both cases, the block should be removed. */
			nlmsvc_unlink_block(block);
		}
	}
	nlmsvc_release_block(block);
}

/* Helper function to handle retry of a deferred block.
 * If it is a blocking lock, call grant_blocked.
 * For a non-blocking lock or test lock, revisit the request.
 */
static void
retry_deferred_block(struct nlm_block *block)
{
	if (!(block->b_flags & B_GOT_CALLBACK))
		block->b_flags |= B_TIMED_OUT;
	nlmsvc_insert_block(block, NLM_TIMEOUT);
	dprintk("revisit block %p flags %d\n",	block, block->b_flags);
	if (block->b_deferred_req) {
		block->b_deferred_req->revisit(block->b_deferred_req, 0);
		block->b_deferred_req = NULL;
	}
}

/*
 * Retry all blocked locks that have been notified. This is where lockd
 * picks up locks that can be granted, or grant notifications that must
 * be retransmitted.
 */
unsigned long
nlmsvc_retry_blocked(void)
{
	unsigned long	timeout = MAX_SCHEDULE_TIMEOUT;
	struct nlm_block *block;

	while (!list_empty(&nlm_blocked) && !kthread_should_stop()) {
		block = list_entry(nlm_blocked.next, struct nlm_block, b_list);

		if (block->b_when == NLM_NEVER)
			break;
		if (time_after(block->b_when, jiffies)) {
			timeout = block->b_when - jiffies;
			break;
		}

		dprintk("nlmsvc_retry_blocked(%p, when=%ld)\n",
			block, block->b_when);
		if (block->b_flags & B_QUEUED) {
			dprintk("nlmsvc_retry_blocked delete block (%p, granted=%d, flags=%d)\n",
				block, block->b_granted, block->b_flags);
			retry_deferred_block(block);
		} else
			nlmsvc_grant_blocked(block);
	}

	return timeout;
}