nfs4state.c

		if (!list_empty(&state->owner->so_seqid.list))
			ret = -EWOULDBLOCK;
	} while (read_seqretry(&state->seqlock, seq));
	return ret;
}

/*
 * Byte-range lock aware utility to initialize the stateid of read/write
 * requests.
 */
int nfs4_select_rw_stateid(nfs4_stateid *dst, struct nfs4_state *state,
		fmode_t fmode, const struct nfs_lockowner *lockowner)
{
	int ret = nfs4_copy_lock_stateid(dst, state, lockowner);
	if (ret == -EIO)
		/* A lost lock - don't even consider delegations */
		goto out;
	if (nfs4_copy_delegation_stateid(dst, state->inode, fmode))
		goto out;
	if (ret != -ENOENT)
		/* nfs4_copy_delegation_stateid() didn't over-write
		 * dst, so it still has the lock stateid which we now
		 * choose to use.
		 */
		goto out;
	ret = nfs4_copy_open_stateid(dst, state);
out:
	if (nfs_server_capable(state->inode, NFS_CAP_STATEID_NFSV41))
		dst->seqid = 0;
	return ret;
}

struct nfs_seqid *nfs_alloc_seqid(struct nfs_seqid_counter *counter, gfp_t gfp_mask)
{
	struct nfs_seqid *new;

	new = kmalloc(sizeof(*new), gfp_mask);
	if (new != NULL) {
		new->sequence = counter;
		INIT_LIST_HEAD(&new->list);
		new->task = NULL;
	}
	return new;
}

void nfs_release_seqid(struct nfs_seqid *seqid)
{
	struct nfs_seqid_counter *sequence;

	if (list_empty(&seqid->list))
		return;
	sequence = seqid->sequence;
	spin_lock(&sequence->lock);
	list_del_init(&seqid->list);
	if (!list_empty(&sequence->list)) {
		struct nfs_seqid *next;

		next = list_first_entry(&sequence->list,
				struct nfs_seqid, list);
		rpc_wake_up_queued_task(&sequence->wait, next->task);
	}
	spin_unlock(&sequence->lock);
}

void nfs_free_seqid(struct nfs_seqid *seqid)
{
	nfs_release_seqid(seqid);
	kfree(seqid);
}

/*
 * Increment the seqid if the OPEN/OPEN_DOWNGRADE/CLOSE succeeded, or
 * failed with a seqid incrementing error -
 * see comments nfs_fs.h:seqid_mutating_error()
 */
static void nfs_increment_seqid(int status, struct nfs_seqid *seqid)
{
	switch (status) {
		case 0:
			break;
		case -NFS4ERR_BAD_SEQID:
			if (seqid->sequence->flags & NFS_SEQID_CONFIRMED)
				return;
			pr_warn_ratelimited("NFS: v4 server returned a bad"
					" sequence-id error on an"
					" unconfirmed sequence %p!\n",
					seqid->sequence);
		case -NFS4ERR_STALE_CLIENTID:
		case -NFS4ERR_STALE_STATEID:
		case -NFS4ERR_BAD_STATEID:
		case -NFS4ERR_BADXDR:
		case -NFS4ERR_RESOURCE:
		case -NFS4ERR_NOFILEHANDLE:
			/* Non-seqid mutating errors */
			return;
	};
	/*
	 * Note: no locking needed as we are guaranteed to be first
	 * on the sequence list
	 */
	seqid->sequence->counter++;
}

void nfs_increment_open_seqid(int status, struct nfs_seqid *seqid)
{
	struct nfs4_state_owner *sp = container_of(seqid->sequence,
					struct nfs4_state_owner, so_seqid);
	struct nfs_server *server = sp->so_server;

	if (status == -NFS4ERR_BAD_SEQID)
		nfs4_drop_state_owner(sp);
	if (!nfs4_has_session(server->nfs_client))
		nfs_increment_seqid(status, seqid);
}

/*
 * Increment the seqid if the LOCK/LOCKU succeeded, or
 * failed with a seqid incrementing error -
 * see comments nfs_fs.h:seqid_mutating_error()
 */
void nfs_increment_lock_seqid(int status, struct nfs_seqid *seqid)
{
	nfs_increment_seqid(status, seqid);
}

int nfs_wait_on_sequence(struct nfs_seqid *seqid, struct rpc_task *task)
{
	struct nfs_seqid_counter *sequence = seqid->sequence;
	int status = 0;

	spin_lock(&sequence->lock);
	seqid->task = task;
	if (list_empty(&seqid->list))
		list_add_tail(&seqid->list, &sequence->list);
	if (list_first_entry(&sequence->list, struct nfs_seqid, list) == seqid)
		goto unlock;
	rpc_sleep_on(&sequence->wait, task, NULL);
	status = -EAGAIN;
unlock:
	spin_unlock(&sequence->lock);
	return status;
}

static int nfs4_run_state_manager(void *);

static void nfs4_clear_state_manager_bit(struct nfs_client *clp)
{
	smp_mb__before_clear_bit();
	clear_bit(NFS4CLNT_MANAGER_RUNNING, &clp->cl_state);
	smp_mb__after_clear_bit();
	wake_up_bit(&clp->cl_state, NFS4CLNT_MANAGER_RUNNING);
	rpc_wake_up(&clp->cl_rpcwaitq);
}

/*
 * Schedule the nfs_client asynchronous state management routine
 */
void nfs4_schedule_state_manager(struct nfs_client *clp)
{
	struct task_struct *task;
	char buf[INET6_ADDRSTRLEN + sizeof("-manager") + 1];

	if (test_and_set_bit(NFS4CLNT_MANAGER_RUNNING, &clp->cl_state) != 0)
		return;
	__module_get(THIS_MODULE);
	atomic_inc(&clp->cl_count);

	/* The rcu_read_lock() is not strictly necessary, as the state
	 * manager is the only thread that ever changes the rpc_xprt
	 * after it's initialized.  At this point, we're single threaded. */
	rcu_read_lock();
	snprintf(buf, sizeof(buf), "%s-manager",
			rpc_peeraddr2str(clp->cl_rpcclient, RPC_DISPLAY_ADDR));
	rcu_read_unlock();
	task = kthread_run(nfs4_run_state_manager, clp, "%s", buf);
	if (IS_ERR(task)) {
		printk(KERN_ERR "%s: kthread_run: %ld\n",
			__func__, PTR_ERR(task));
		nfs4_clear_state_manager_bit(clp);
		nfs_put_client(clp);
		module_put(THIS_MODULE);
	}
}

/*
 * Schedule a lease recovery attempt
 */
void nfs4_schedule_lease_recovery(struct nfs_client *clp)
{
	if (!clp)
		return;
	if (!test_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state))
		set_bit(NFS4CLNT_CHECK_LEASE, &clp->cl_state);
	dprintk("%s: scheduling lease recovery for server %s\n", __func__,
			clp->cl_hostname);
	nfs4_schedule_state_manager(clp);
}
EXPORT_SYMBOL_GPL(nfs4_schedule_lease_recovery);

/**
 * nfs4_schedule_migration_recovery - trigger migration recovery
 *
 * @server: FSID that is migrating
 *
 * Returns zero if recovery has started, otherwise a negative NFS4ERR
 * value is returned.
 */
int nfs4_schedule_migration_recovery(const struct nfs_server *server)
{
	struct nfs_client *clp = server->nfs_client;

	if (server->fh_expire_type != NFS4_FH_PERSISTENT) {
		pr_err("NFS: volatile file handles not supported (server %s)\n",
				clp->cl_hostname);
		return -NFS4ERR_IO;
	}

	if (test_bit(NFS_MIG_FAILED, &server->mig_status))
		return -NFS4ERR_IO;

	dprintk("%s: scheduling migration recovery for (%llx:%llx) on %s\n",
			__func__,
			(unsigned long long)server->fsid.major,
			(unsigned long long)server->fsid.minor,
			clp->cl_hostname);

	set_bit(NFS_MIG_IN_TRANSITION,
			&((struct nfs_server *)server)->mig_status);
	set_bit(NFS4CLNT_MOVED, &clp->cl_state);

	nfs4_schedule_state_manager(clp);
	return 0;
}
EXPORT_SYMBOL_GPL(nfs4_schedule_migration_recovery);

/**
 * nfs4_schedule_lease_moved_recovery - start lease-moved recovery
 *
 * @clp: server to check for moved leases
 *
 */
void nfs4_schedule_lease_moved_recovery(struct nfs_client *clp)
{
	dprintk("%s: scheduling lease-moved recovery for client ID %llx on %s\n",
		__func__, clp->cl_clientid, clp->cl_hostname);

	set_bit(NFS4CLNT_LEASE_MOVED, &clp->cl_state);
	nfs4_schedule_state_manager(clp);
}
EXPORT_SYMBOL_GPL(nfs4_schedule_lease_moved_recovery);

int nfs4_wait_clnt_recover(struct nfs_client *clp)
{
	int res;

	might_sleep();

	atomic_inc(&clp->cl_count);
	res = wait_on_bit(&clp->cl_state, NFS4CLNT_MANAGER_RUNNING,
			nfs_wait_bit_killable, TASK_KILLABLE);
	if (res)
		goto out;
	if (clp->cl_cons_state < 0)
		res = clp->cl_cons_state;
out:
	nfs_put_client(clp);
	return res;
}

int nfs4_client_recover_expired_lease(struct nfs_client *clp)
{
	unsigned int loop;
	int ret;

	for (loop = NFS4_MAX_LOOP_ON_RECOVER; loop != 0; loop--) {
		ret = nfs4_wait_clnt_recover(clp);
		if (ret != 0)
			break;
		if (!test_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state) &&
		    !test_bit(NFS4CLNT_CHECK_LEASE,&clp->cl_state))
			break;
		nfs4_schedule_state_manager(clp);
		ret = -EIO;
	}
	return ret;
}

/*
 * nfs40_handle_cb_pathdown - return all delegations after NFS4ERR_CB_PATH_DOWN
 * @clp: client to process
 *
 * Set the NFS4CLNT_LEASE_EXPIRED state in order to force a
 * resend of the SETCLIENTID and hence re-establish the
 * callback channel. Then return all existing delegations.
 */
static void nfs40_handle_cb_pathdown(struct nfs_client *clp)
{
	set_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state);
	nfs_expire_all_delegations(clp);
	dprintk("%s: handling CB_PATHDOWN recovery for server %s\n", __func__,
			clp->cl_hostname);
}

void nfs4_schedule_path_down_recovery(struct nfs_client *clp)
{
	nfs40_handle_cb_pathdown(clp);
	nfs4_schedule_state_manager(clp);
}

static int nfs4_state_mark_reclaim_reboot(struct nfs_client *clp, struct nfs4_state *state)
{

	set_bit(NFS_STATE_RECLAIM_REBOOT, &state->flags);
	/* Don't recover state that expired before the reboot */
	if (test_bit(NFS_STATE_RECLAIM_NOGRACE, &state->flags)) {
		clear_bit(NFS_STATE_RECLAIM_REBOOT, &state->flags);
		return 0;
	}
	set_bit(NFS_OWNER_RECLAIM_REBOOT, &state->owner->so_flags);
	set_bit(NFS4CLNT_RECLAIM_REBOOT, &clp->cl_state);
	return 1;
}

static int nfs4_state_mark_reclaim_nograce(struct nfs_client *clp, struct nfs4_state *state)
{
	set_bit(NFS_STATE_RECLAIM_NOGRACE, &state->flags);
	clear_bit(NFS_STATE_RECLAIM_REBOOT, &state->flags);
	set_bit(NFS_OWNER_RECLAIM_NOGRACE, &state->owner->so_flags);
	set_bit(NFS4CLNT_RECLAIM_NOGRACE, &clp->cl_state);
	return 1;
}

int nfs4_schedule_stateid_recovery(const struct nfs_server *server, struct nfs4_state *state)
{
	struct nfs_client *clp = server->nfs_client;

	if (!nfs4_valid_open_stateid(state))
		return -EBADF;
	nfs4_state_mark_reclaim_nograce(clp, state);
	dprintk("%s: scheduling stateid recovery for server %s\n", __func__,
			clp->cl_hostname);
	nfs4_schedule_state_manager(clp);
	return 0;
}
EXPORT_SYMBOL_GPL(nfs4_schedule_stateid_recovery);

void nfs_inode_find_state_and_recover(struct inode *inode,
		const nfs4_stateid *stateid)
{
	struct nfs_client *clp = NFS_SERVER(inode)->nfs_client;
	struct nfs_inode *nfsi = NFS_I(inode);
	struct nfs_open_context *ctx;
	struct nfs4_state *state;
	bool found = false;

	spin_lock(&inode->i_lock);
	list_for_each_entry(ctx, &nfsi->open_files, list) {
		state = ctx->state;
		if (state == NULL)
			continue;
		if (!test_bit(NFS_DELEGATED_STATE, &state->flags))
			continue;
		if (!nfs4_stateid_match(&state->stateid, stateid))
			continue;
		nfs4_state_mark_reclaim_nograce(clp, state);
		found = true;
	}
	spin_unlock(&inode->i_lock);
	if (found)
		nfs4_schedule_state_manager(clp);
}

static void nfs4_state_mark_open_context_bad(struct nfs4_state *state)
{
	struct inode *inode = state->inode;
	struct nfs_inode *nfsi = NFS_I(inode);
	struct nfs_open_context *ctx;

	spin_lock(&inode->i_lock);
	list_for_each_entry(ctx, &nfsi->open_files, list) {
		if (ctx->state != state)
			continue;
		set_bit(NFS_CONTEXT_BAD, &ctx->flags);
	}
	spin_unlock(&inode->i_lock);
}

static void nfs4_state_mark_recovery_failed(struct nfs4_state *state, int error)
{
	set_bit(NFS_STATE_RECOVERY_FAILED, &state->flags);
	nfs4_state_mark_open_context_bad(state);
}


static int nfs4_reclaim_locks(struct nfs4_state *state, const struct nfs4_state_recovery_ops *ops)
{
	struct inode *inode = state->inode;
	struct nfs_inode *nfsi = NFS_I(inode);
	struct file_lock *fl;
	int status = 0;

	if (inode->i_flock == NULL)
		return 0;

	/* Guard against delegation returns and new lock/unlock calls */
	down_write(&nfsi->rwsem);
	/* Protect inode->i_flock using the BKL */
	spin_lock(&inode->i_lock);
	for (fl = inode->i_flock; fl != NULL; fl = fl->fl_next) {
		if (!(fl->fl_flags & (FL_POSIX|FL_FLOCK)))
			continue;
		if (nfs_file_open_context(fl->fl_file)->state != state)
			continue;
		spin_unlock(&inode->i_lock);
		status = ops->recover_lock(state, fl);
		switch (status) {
			case 0:
				break;
			case -ESTALE:
			case -NFS4ERR_ADMIN_REVOKED:
			case -NFS4ERR_STALE_STATEID:
			case -NFS4ERR_BAD_STATEID:
			case -NFS4ERR_EXPIRED:
			case -NFS4ERR_NO_GRACE:
			case -NFS4ERR_STALE_CLIENTID:
			case -NFS4ERR_BADSESSION:
			case -NFS4ERR_BADSLOT:
			case -NFS4ERR_BAD_HIGH_SLOT:
			case -NFS4ERR_CONN_NOT_BOUND_TO_SESSION:
				goto out;
			default:
				printk(KERN_ERR "NFS: %s: unhandled error %d\n",
					 __func__, status);
			case -ENOMEM:
			case -NFS4ERR_DENIED:
			case -NFS4ERR_RECLAIM_BAD:
			case -NFS4ERR_RECLAIM_CONFLICT:
				/* kill_proc(fl->fl_pid, SIGLOST, 1); */
				status = 0;
		}
		spin_lock(&inode->i_lock);
	}
	spin_unlock(&inode->i_lock);
out:
	up_write(&nfsi->rwsem);
	return status;
}

static int nfs4_reclaim_open_state(struct nfs4_state_owner *sp, const struct nfs4_state_recovery_ops *ops)
{
	struct nfs4_state *state;
	struct nfs4_lock_state *lock;
	int status = 0;

	/* Note: we rely on the sp->so_states list being ordered 
	 * so that we always reclaim open(O_RDWR) and/or open(O_WRITE)
	 * states first.
	 * This is needed to ensure that the server won't give us any
	 * read delegations that we have to return if, say, we are
	 * recovering after a network partition or a reboot from a
	 * server that doesn't support a grace period.
	 */
	spin_lock(&sp->so_lock);
	write_seqcount_begin(&sp->so_reclaim_seqcount);
restart:
	list_for_each_entry(state, &sp->so_states, open_states) {
		if (!test_and_clear_bit(ops->state_flag_bit, &state->flags))
			continue;
		if (!nfs4_valid_open_stateid(state))
			continue;
		if (state->state == 0)
			continue;
		atomic_inc(&state->count);
		spin_unlock(&sp->so_lock);
		status = ops->recover_open(sp, state);
		if (status >= 0) {
			status = nfs4_reclaim_locks(state, ops);
			if (status >= 0) {
				if (!test_bit(NFS_DELEGATED_STATE, &state->flags)) {
					spin_lock(&state->state_lock);
					list_for_each_entry(lock, &state->lock_states, ls_locks) {
						if (!test_bit(NFS_LOCK_INITIALIZED, &lock->ls_flags))
							pr_warn_ratelimited("NFS: "
									    "%s: Lock reclaim "
									    "failed!\n", __func__);
					}
					spin_unlock(&state->state_lock);
				}
				nfs4_put_open_state(state);
				spin_lock(&sp->so_lock);
				goto restart;
			}
		}
		switch (status) {
			default:
				printk(KERN_ERR "NFS: %s: unhandled error %d\n",
					__func__, status);
			case -ENOENT:
			case -ENOMEM:
			case -ESTALE:
				/* Open state on this file cannot be recovered */
				nfs4_state_mark_recovery_failed(state, status);
				break;
			case -EAGAIN:
				ssleep(1);
			case -NFS4ERR_ADMIN_REVOKED:
			case -NFS4ERR_STALE_STATEID:
			case -NFS4ERR_BAD_STATEID:
			case -NFS4ERR_RECLAIM_BAD:
			case -NFS4ERR_RECLAIM_CONFLICT:
				nfs4_state_mark_reclaim_nograce(sp->so_server->nfs_client, state);
				break;
			case -NFS4ERR_EXPIRED:
			case -NFS4ERR_NO_GRACE:
				nfs4_state_mark_reclaim_nograce(sp->so_server->nfs_client, state);
			case -NFS4ERR_STALE_CLIENTID:
			case -NFS4ERR_BADSESSION:
			case -NFS4ERR_BADSLOT:
			case -NFS4ERR_BAD_HIGH_SLOT:
			case -NFS4ERR_CONN_NOT_BOUND_TO_SESSION:
				goto out_err;
		}
		nfs4_put_open_state(state);
		spin_lock(&sp->so_lock);
		goto restart;
	}
	write_seqcount_end(&sp->so_reclaim_seqcount);
	spin_unlock(&sp->so_lock);
	return 0;
out_err:
	nfs4_put_open_state(state);
	spin_lock(&sp->so_lock);
	write_seqcount_end(&sp->so_reclaim_seqcount);
	spin_unlock(&sp->so_lock);
	return status;
}

static void nfs4_clear_open_state(struct nfs4_state *state)
{
	struct nfs4_lock_state *lock;

	clear_bit(NFS_DELEGATED_STATE, &state->flags);
	clear_bit(NFS_O_RDONLY_STATE, &state->flags);
	clear_bit(NFS_O_WRONLY_STATE, &state->flags);
	clear_bit(NFS_O_RDWR_STATE, &state->flags);
	spin_lock(&state->state_lock);
	list_for_each_entry(lock, &state->lock_states, ls_locks) {
		lock->ls_seqid.flags = 0;
		clear_bit(NFS_LOCK_INITIALIZED, &lock->ls_flags);
	}
	spin_unlock(&state->state_lock);
}

static void nfs4_reset_seqids(struct nfs_server *server,
	int (*mark_reclaim)(struct nfs_client *clp, struct nfs4_state *state))
{
	struct nfs_client *clp = server->nfs_client;
	struct nfs4_state_owner *sp;
	struct rb_node *pos;
	struct nfs4_state *state;

	spin_lock(&clp->cl_lock);
	for (pos = rb_first(&server->state_owners);
	     pos != NULL;
	     pos = rb_next(pos)) {
		sp = rb_entry(pos, struct nfs4_state_owner, so_server_node);
		sp->so_seqid.flags = 0;
		spin_lock(&sp->so_lock);
		list_for_each_entry(state, &sp->so_states, open_states) {
			if (mark_reclaim(clp, state))
				nfs4_clear_open_state(state);
		}
		spin_unlock(&sp->so_lock);
	}
	spin_unlock(&clp->cl_lock);
}

static void nfs4_state_mark_reclaim_helper(struct nfs_client *clp,
	int (*mark_reclaim)(struct nfs_client *clp, struct nfs4_state *state))
{
	struct nfs_server *server;

	rcu_read_lock();
	list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link)
		nfs4_reset_seqids(server, mark_reclaim);
	rcu_read_unlock();
}

static void nfs4_state_start_reclaim_reboot(struct nfs_client *clp)
{
	/* Mark all delegations for reclaim */
	nfs_delegation_mark_reclaim(clp);
	nfs4_state_mark_reclaim_helper(clp, nfs4_state_mark_reclaim_reboot);
}

static void nfs4_reclaim_complete(struct nfs_client *clp,
				 const struct nfs4_state_recovery_ops *ops,
				 struct rpc_cred *cred)
{
	/* Notify the server we're done reclaiming our state */
	if (ops->reclaim_complete)
		(void)ops->reclaim_complete(clp, cred);
}

static void nfs4_clear_reclaim_server(struct nfs_server *server)
{
	struct nfs_client *clp = server->nfs_client;
	struct nfs4_state_owner *sp;
	struct rb_node *pos;
	struct nfs4_state *state;

	spin_lock(&clp->cl_lock);
	for (pos = rb_first(&server->state_owners);
	     pos != NULL;
	     pos = rb_next(pos)) {
		sp = rb_entry(pos, struct nfs4_state_owner, so_server_node);
		spin_lock(&sp->so_lock);
		list_for_each_entry(state, &sp->so_states, open_states) {
			if (!test_and_clear_bit(NFS_STATE_RECLAIM_REBOOT,
						&state->flags))
				continue;
			nfs4_state_mark_reclaim_nograce(clp, state);
		}
		spin_unlock(&sp->so_lock);
	}
	spin_unlock(&clp->cl_lock);
}

static int nfs4_state_clear_reclaim_reboot(struct nfs_client *clp)
{
	struct nfs_server *server;

	if (!test_and_clear_bit(NFS4CLNT_RECLAIM_REBOOT, &clp->cl_state))
		return 0;

	rcu_read_lock();
	list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link)
		nfs4_clear_reclaim_server(server);
	rcu_read_unlock();

	nfs_delegation_reap_unclaimed(clp);
	return 1;
}

static void nfs4_state_end_reclaim_reboot(struct nfs_client *clp)
{
	const struct nfs4_state_recovery_ops *ops;
	struct rpc_cred *cred;

	if (!nfs4_state_clear_reclaim_reboot(clp))
		return;
	ops = clp->cl_mvops->reboot_recovery_ops;
	cred = nfs4_get_clid_cred(clp);
	nfs4_reclaim_complete(clp, ops, cred);
	put_rpccred(cred);
}

static void nfs_delegation_clear_all(struct nfs_client *clp)
{
	nfs_delegation_mark_reclaim(clp);
	nfs_delegation_reap_unclaimed(clp);
}

static void nfs4_state_start_reclaim_nograce(struct nfs_client *clp)
{
	nfs_delegation_clear_all(clp);
	nfs4_state_mark_reclaim_helper(clp, nfs4_state_mark_reclaim_nograce);
}

static int nfs4_recovery_handle_error(struct nfs_client *clp, int error)
{
	switch (error) {
		case 0:
			break;
		case -NFS4ERR_CB_PATH_DOWN:
			nfs40_handle_cb_pathdown(clp);
			break;
		case -NFS4ERR_NO_GRACE:
			nfs4_state_end_reclaim_reboot(clp);
			break;
		case -NFS4ERR_STALE_CLIENTID:
			set_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state);
			nfs4_state_clear_reclaim_reboot(clp);
			nfs4_state_start_reclaim_reboot(clp);
			break;
		case -NFS4ERR_EXPIRED:
			set_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state);
			nfs4_state_start_reclaim_nograce(clp);
			break;
		case -NFS4ERR_BADSESSION:
		case -NFS4ERR_BADSLOT:
		case -NFS4ERR_BAD_HIGH_SLOT:
		case -NFS4ERR_DEADSESSION:
		case -NFS4ERR_SEQ_FALSE_RETRY:
		case -NFS4ERR_SEQ_MISORDERED:
			set_bit(NFS4CLNT_SESSION_RESET, &clp->cl_state);
			/* Zero session reset errors */
			break;
		case -NFS4ERR_CONN_NOT_BOUND_TO_SESSION:
			set_bit(NFS4CLNT_BIND_CONN_TO_SESSION, &clp->cl_state);
			break;
		default:
			dprintk("%s: failed to handle error %d for server %s\n",
					__func__, error, clp->cl_hostname);
			return error;
	}
	dprintk("%s: handled error %d for server %s\n", __func__, error,
			clp->cl_hostname);
	return 0;
}

static int nfs4_do_reclaim(struct nfs_client *clp, const struct nfs4_state_recovery_ops *ops)
{
	struct nfs4_state_owner *sp;
	struct nfs_server *server;
	struct rb_node *pos;
	int status = 0;

restart:
	rcu_read_lock();
	list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link) {
		nfs4_purge_state_owners(server);
		spin_lock(&clp->cl_lock);
		for (pos = rb_first(&server->state_owners);
		     pos != NULL;
		     pos = rb_next(pos)) {
			sp = rb_entry(pos,
				struct nfs4_state_owner, so_server_node);
			if (!test_and_clear_bit(ops->owner_flag_bit,
							&sp->so_flags))
				continue;
			atomic_inc(&sp->so_count);
			spin_unlock(&clp->cl_lock);
			rcu_read_unlock();

			status = nfs4_reclaim_open_state(sp, ops);
			if (status < 0) {
				set_bit(ops->owner_flag_bit, &sp->so_flags);
				nfs4_put_state_owner(sp);
				return nfs4_recovery_handle_error(clp, status);
			}

			nfs4_put_state_owner(sp);
			goto restart;
		}
		spin_unlock(&clp->cl_lock);
	}
	rcu_read_unlock();
	return status;
}

static int nfs4_check_lease(struct nfs_client *clp)
{
	struct rpc_cred *cred;
	const struct nfs4_state_maintenance_ops *ops =
		clp->cl_mvops->state_renewal_ops;
	int status;

	/* Is the client already known to have an expired lease? */
	if (test_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state))
		return 0;
	spin_lock(&clp->cl_lock);
	cred = ops->get_state_renewal_cred_locked(clp);
	spin_unlock(&clp->cl_lock);
	if (cred == NULL) {
		cred = nfs4_get_clid_cred(clp);
		status = -ENOKEY;
		if (cred == NULL)
			goto out;
	}
	status = ops->renew_lease(clp, cred);
	put_rpccred(cred);
	if (status == -ETIMEDOUT) {
		set_bit(NFS4CLNT_CHECK_LEASE, &clp->cl_state);
		return 0;
	}
out:
	return nfs4_recovery_handle_error(clp, status);
}

/* Set NFS4CLNT_LEASE_EXPIRED and reclaim reboot state for all v4.0 errors
 * and for recoverable errors on EXCHANGE_ID for v4.1
 */
static int nfs4_handle_reclaim_lease_error(struct nfs_client *clp, int status)
{
	switch (status) {
	case -NFS4ERR_SEQ_MISORDERED:
		if (test_and_set_bit(NFS4CLNT_PURGE_STATE, &clp->cl_state))
			return -ESERVERFAULT;
		/* Lease confirmation error: retry after purging the lease */
		ssleep(1);
		clear_bit(NFS4CLNT_LEASE_CONFIRM, &clp->cl_state);
		break;
	case -NFS4ERR_STALE_CLIENTID:
		clear_bit(NFS4CLNT_LEASE_CONFIRM, &clp->cl_state);
		nfs4_state_clear_reclaim_reboot(clp);
		nfs4_state_start_reclaim_reboot(clp);
		break;
	case -NFS4ERR_CLID_INUSE:
		pr_err("NFS: Server %s reports our clientid is in use\n",
			clp->cl_hostname);
		nfs_mark_client_ready(clp, -EPERM);
		clear_bit(NFS4CLNT_LEASE_CONFIRM, &clp->cl_state);
		return -EPERM;
	case -EACCES:
	case -NFS4ERR_DELAY:
	case -ETIMEDOUT:
	case -EAGAIN:
		ssleep(1);
		break;

	case -NFS4ERR_MINOR_VERS_MISMATCH:
		if (clp->cl_cons_state == NFS_CS_SESSION_INITING)
			nfs_mark_client_ready(clp, -EPROTONOSUPPORT);
		dprintk("%s: exit with error %d for server %s\n",
				__func__, -EPROTONOSUPPORT, clp->cl_hostname);
		return -EPROTONOSUPPORT;
	case -NFS4ERR_NOT_SAME: /* FixMe: implement recovery
				 * in nfs4_exchange_id */
	default:
		dprintk("%s: exit with error %d for server %s\n", __func__,
				status, clp->cl_hostname);
		return status;
	}
	set_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state);
	dprintk("%s: handled error %d for server %s\n", __func__, status,
			clp->cl_hostname);
	return 0;
}

static int nfs4_establish_lease(struct nfs_client *clp)
{
	struct rpc_cred *cred;
	const struct nfs4_state_recovery_ops *ops =
		clp->cl_mvops->reboot_recovery_ops;
	int status;

	cred = nfs4_get_clid_cred(clp);
	if (cred == NULL)
		return -ENOENT;
	status = ops->establish_clid(clp, cred);
	put_rpccred(cred);
	if (status != 0)
		return status;
	pnfs_destroy_all_layouts(clp);
	return 0;
}

/*
 * Returns zero or a negative errno.  NFS4ERR values are converted
 * to local errno values.
 */
static int nfs4_reclaim_lease(struct nfs_client *clp)
{
	int status;

	status = nfs4_establish_lease(clp);
	if (status < 0)
		return nfs4_handle_reclaim_lease_error(clp, status);
	if (test_and_clear_bit(NFS4CLNT_SERVER_SCOPE_MISMATCH, &clp->cl_state))
		nfs4_state_start_reclaim_nograce(clp);
	if (!test_bit(NFS4CLNT_RECLAIM_NOGRACE, &clp->cl_state))
		set_bit(NFS4CLNT_RECLAIM_REBOOT, &clp->cl_state);
	clear_bit(NFS4CLNT_CHECK_LEASE, &clp->cl_state);
	clear_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state);
	return 0;
}

static int nfs4_purge_lease(struct nfs_client *clp)
{
	int status;

	status = nfs4_establish_lease(clp);
	if (status < 0)
		return nfs4_handle_reclaim_lease_error(clp, status);
	clear_bit(NFS4CLNT_PURGE_STATE, &clp->cl_state);
	set_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state);
	nfs4_state_start_reclaim_nograce(clp);
	return 0;
}

/*
 * Try remote migration of one FSID from a source server to a
 * destination server.  The source server provides a list of
 * potential destinations.
 *
 * Returns zero or a negative NFS4ERR status code.
 */
static int nfs4_try_migration(struct nfs_server *server, struct rpc_cred *cred)
{
	struct nfs_client *clp = server->nfs_client;
	struct nfs4_fs_locations *locations = NULL;
	struct inode *inode;
	struct page *page;
	int status, result;

	dprintk("--> %s: FSID %llx:%llx on \"%s\"\n", __func__,
			(unsigned long long)server->fsid.major,
			(unsigned long long)server->fsid.minor,
			clp->cl_hostname);

	result = 0;
	page = alloc_page(GFP_KERNEL);
	locations = kmalloc(sizeof(struct nfs4_fs_locations), GFP_KERNEL);
	if (page == NULL || locations == NULL) {
		dprintk("<-- %s: no memory\n", __func__);
		goto out;
	}

	inode = server->super->s_root->d_inode;
	result = nfs4_proc_get_locations(inode, locations, page, cred);
	if (result) {
		dprintk("<-- %s: failed to retrieve fs_locations: %d\n",
			__func__, result);
		goto out;
	}

	result = -NFS4ERR_NXIO;
	if (!(locations->fattr.valid & NFS_ATTR_FATTR_V4_LOCATIONS)) {
		dprintk("<-- %s: No fs_locations data, migration skipped\n",
			__func__);
		goto out;
	}

	nfs4_begin_drain_session(clp);

	status = nfs4_replace_transport(server, locations);
	if (status != 0) {
		dprintk("<-- %s: failed to replace transport: %d\n",
			__func__, status);
		goto out;
	}

	result = 0;
	dprintk("<-- %s: migration succeeded\n", __func__);

out:
	if (page != NULL)
		__free_page(page);
	kfree(locations);
	if (result) {
		pr_err("NFS: migration recovery failed (server %s)\n",
				clp->cl_hostname);
		set_bit(NFS_MIG_FAILED, &server->mig_status);
	}
	return result;
}

/*
 * Returns zero or a negative NFS4ERR status code.
 */
static int nfs4_handle_migration(struct nfs_client *clp)
{
	const struct nfs4_state_maintenance_ops *ops =
				clp->cl_mvops->state_renewal_ops;
	struct nfs_server *server;
	struct rpc_cred *cred;

	dprintk("%s: migration reported on \"%s\"\n", __func__,
			clp->cl_hostname);

	spin_lock(&clp->cl_lock);
	cred = ops->get_state_renewal_cred_locked(clp);
	spin_unlock(&clp->cl_lock);
	if (cred == NULL)
		return -NFS4ERR_NOENT;

	clp->cl_mig_gen++;
restart:
	rcu_read_lock();
	list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link) {
		int status;

		if (server->mig_gen == clp->cl_mig_gen)
			continue;
		server->mig_gen = clp->cl_mig_gen;

		if (!test_and_clear_bit(NFS_MIG_IN_TRANSITION,
						&server->mig_status))
			continue;

		rcu_read_unlock();
		status = nfs4_try_migration(server, cred);
		if (status < 0) {
			put_rpccred(cred);
			return status;
		}
		goto restart;
	}
	rcu_read_unlock();
	put_rpccred(cred);
	return 0;
}

/*
 * Test each nfs_server on the clp's cl_superblocks list to see
 * if it's moved to another server.  Stop when the server no longer
 * returns NFS4ERR_LEASE_MOVED.
 */
static int nfs4_handle_lease_moved(struct nfs_client *clp)