Merge tag 'nfs-for-3.6-1' of git://git.linux-nfs.org/projects/trondmy/linux-nfs

config NFS_V4_1

	bool "NFS client support for NFSv4.1 (EXPERIMENTAL)"

	depends on NFS_FS && NFS_V4 && EXPERIMENTAL

	depends on NFS_V4 && EXPERIMENTAL

	select SUNRPC_BACKCHANNEL

	select PNFS_FILE_LAYOUT

	help

	  This option enables support for minor version 1 of the NFSv4 protocol

	  (RFC 5661) in the kernel's NFS client.

config PNFS_FILE_LAYOUT

	tristate

	depends on NFS_V4_1

	default m

config PNFS_BLOCK

	tristate

	depends on NFS_FS && NFS_V4_1 && BLK_DEV_DM

	depends on NFS_V4_1 && BLK_DEV_DM

	default m

config PNFS_OBJLAYOUT

	tristate

	depends on NFS_FS && NFS_V4_1 && SCSI_OSD_ULD

	depends on NFS_V4_1 && SCSI_OSD_ULD

	default m

config NFS_V4_1_IMPLEMENTATION_ID_DOMAIN

nfs-$(CONFIG_NFS_V3)	+= nfs3proc.o nfs3xdr.o

nfs-$(CONFIG_NFS_V3_ACL)	+= nfs3acl.o

nfs-$(CONFIG_NFS_V4)	+= nfs4proc.o nfs4xdr.o nfs4state.o nfs4renewd.o \

			   delegation.o idmap.o \

			   nfs4super.o nfs4file.o delegation.o idmap.o \

			   callback.o callback_xdr.o callback_proc.o \

			   nfs4namespace.o

			   nfs4namespace.o nfs4getroot.o nfs4client.o

nfs-$(CONFIG_NFS_V4_1)	+= pnfs.o pnfs_dev.o

nfs-$(CONFIG_SYSCTL) += sysctl.o

ifeq ($(CONFIG_SYSCTL), y)

nfs-y += sysctl.o

nfs-$(CONFIG_NFS_V4) += nfs4sysctl.o

endif

nfs-$(CONFIG_NFS_FSCACHE) += fscache.o fscache-index.o

obj-$(CONFIG_PNFS_FILE_LAYOUT) += nfs_layout_nfsv41_files.o

#define NFSDBG_FACILITY		NFSDBG_CLIENT

static DECLARE_WAIT_QUEUE_HEAD(nfs_client_active_wq);

#ifdef CONFIG_NFS_V4

/*

 * Get a unique NFSv4.0 callback identifier which will be used

 * by the V4.0 callback service to lookup the nfs_client struct

 */

static int nfs_get_cb_ident_idr(struct nfs_client *clp, int minorversion)

{

	int ret = 0;

	struct nfs_net *nn = net_generic(clp->cl_net, nfs_net_id);

	if (clp->rpc_ops->version != 4 || minorversion != 0)

		return ret;

retry:

	if (!idr_pre_get(&nn->cb_ident_idr, GFP_KERNEL))

		return -ENOMEM;

	spin_lock(&nn->nfs_client_lock);

	ret = idr_get_new(&nn->cb_ident_idr, clp, &clp->cl_cb_ident);

	spin_unlock(&nn->nfs_client_lock);

	if (ret == -EAGAIN)

		goto retry;

	return ret;

}

#endif /* CONFIG_NFS_V4 */

/*

 * Turn off NFSv4 uid/gid mapping when using AUTH_SYS

 */

static bool nfs4_disable_idmapping = true;

/*

 * RPC cruft for NFS

};

#endif  /* CONFIG_NFS_V3_ACL */

struct nfs_client_initdata {

	unsigned long init_flags;

	const char *hostname;

	const struct sockaddr *addr;

	size_t addrlen;

	const struct nfs_rpc_ops *rpc_ops;

	int proto;

	u32 minorversion;

	struct net *net;

};

/*

 * Allocate a shared client record

 *

 * Since these are allocated/deallocated very rarely, we don't

 * bother putting them in a slab cache...

 */

static struct nfs_client *nfs_alloc_client(const struct nfs_client_initdata *cl_init)

struct nfs_client *nfs_alloc_client(const struct nfs_client_initdata *cl_init)

{

	struct nfs_client *clp;

	struct rpc_cred *cred;

	clp->cl_proto = cl_init->proto;

	clp->cl_net = get_net(cl_init->net);

#ifdef CONFIG_NFS_V4

	err = nfs_get_cb_ident_idr(clp, cl_init->minorversion);

	if (err)

		goto error_cleanup;

	spin_lock_init(&clp->cl_lock);

	INIT_DELAYED_WORK(&clp->cl_renewd, nfs4_renew_state);

	rpc_init_wait_queue(&clp->cl_rpcwaitq, "NFS client");

	clp->cl_state = 1 << NFS4CLNT_LEASE_EXPIRED;

	clp->cl_minorversion = cl_init->minorversion;

	clp->cl_mvops = nfs_v4_minor_ops[cl_init->minorversion];

#endif

	cred = rpc_lookup_machine_cred("*");

	if (!IS_ERR(cred))

		clp->cl_machine_cred = cred;

}

#ifdef CONFIG_NFS_V4

#ifdef CONFIG_NFS_V4_1

static void nfs4_shutdown_session(struct nfs_client *clp)

{

	if (nfs4_has_session(clp)) {

		nfs4_destroy_session(clp->cl_session);

		nfs4_destroy_clientid(clp);

	}

}

#else /* CONFIG_NFS_V4_1 */

static void nfs4_shutdown_session(struct nfs_client *clp)

{

}

#endif /* CONFIG_NFS_V4_1 */

/*

 * Destroy the NFS4 callback service

 */

static void nfs4_destroy_callback(struct nfs_client *clp)

{

	if (__test_and_clear_bit(NFS_CS_CALLBACK, &clp->cl_res_state))

		nfs_callback_down(clp->cl_mvops->minor_version);

}

static void nfs4_shutdown_client(struct nfs_client *clp)

{

	if (__test_and_clear_bit(NFS_CS_RENEWD, &clp->cl_res_state))

		nfs4_kill_renewd(clp);

	nfs4_shutdown_session(clp);

	nfs4_destroy_callback(clp);

	if (__test_and_clear_bit(NFS_CS_IDMAP, &clp->cl_res_state))

		nfs_idmap_delete(clp);

	rpc_destroy_wait_queue(&clp->cl_rpcwaitq);

	kfree(clp->cl_serverowner);

	kfree(clp->cl_serverscope);

	kfree(clp->cl_implid);

}

/* idr_remove_all is not needed as all id's are removed by nfs_put_client */

void nfs_cleanup_cb_ident_idr(struct net *net)

{

	rpc_init_wait_queue(&server->roc_rpcwaitq, "pNFS ROC");

}

static void nfs4_destroy_server(struct nfs_server *server)

{

	nfs4_purge_state_owners(server);

}

#else

static void nfs4_shutdown_client(struct nfs_client *clp)

{

}

void nfs_cleanup_cb_ident_idr(struct net *net)

{

}

/*

 * Destroy a shared client record

 */

static void nfs_free_client(struct nfs_client *clp)

void nfs_free_client(struct nfs_client *clp)

{

	dprintk("--> nfs_free_client(%u)\n", clp->rpc_ops->version);

	nfs4_shutdown_client(clp);

	nfs_fscache_release_client_cookie(clp);

	/* -EIO all pending I/O */

		BUG_ON(!list_empty(&clp->cl_superblocks));

		nfs_free_client(clp);

		clp->rpc_ops->free_client(clp);

	}

}

EXPORT_SYMBOL_GPL(nfs_put_client);

 * Test if two socket addresses represent the same actual socket,

 * by comparing (only) relevant fields, excluding the port number.

 */

static int nfs_sockaddr_match_ipaddr(const struct sockaddr *sa1,

int nfs_sockaddr_match_ipaddr(const struct sockaddr *sa1,

			      const struct sockaddr *sa2)

{

	if (sa1->sa_family != sa2->sa_family)

	return 0;

}

#if defined(CONFIG_NFS_V4_1)

/* Common match routine for v4.0 and v4.1 callback services */

static bool nfs4_cb_match_client(const struct sockaddr *addr,

		struct nfs_client *clp, u32 minorversion)

{

	struct sockaddr *clap = (struct sockaddr *)&clp->cl_addr;

	/* Don't match clients that failed to initialise */

	if (!(clp->cl_cons_state == NFS_CS_READY ||

	    clp->cl_cons_state == NFS_CS_SESSION_INITING))

		return false;

	smp_rmb();

	/* Match the version and minorversion */

	if (clp->rpc_ops->version != 4 ||

	    clp->cl_minorversion != minorversion)

		return false;

	/* Match only the IP address, not the port number */

	if (!nfs_sockaddr_match_ipaddr(addr, clap))

		return false;

	return true;

}

#endif /* CONFIG_NFS_V4_1 */

/*

 * Find an nfs_client on the list that matches the initialisation data

 * that is supplied.

 * Look up a client by IP address and protocol version

 * - creates a new record if one doesn't yet exist

 */

static struct nfs_client *

struct nfs_client *

nfs_get_client(const struct nfs_client_initdata *cl_init,

	       const struct rpc_timeout *timeparms,

	       const char *ip_addr,

		if (clp) {

			spin_unlock(&nn->nfs_client_lock);

			if (new)

				nfs_free_client(new);

				new->rpc_ops->free_client(new);

			return nfs_found_client(cl_init, clp);

		}

		if (new) {

		spin_unlock(&nn->nfs_client_lock);

		new = nfs_alloc_client(cl_init);

		new = cl_init->rpc_ops->alloc_client(cl_init);

	} while (!IS_ERR(new));

	dprintk("<-- nfs_get_client() Failed to find %s (%ld)\n",

/*

 * Initialise the timeout values for a connection

 */

static void nfs_init_timeout_values(struct rpc_timeout *to, int proto,

void nfs_init_timeout_values(struct rpc_timeout *to, int proto,

				    unsigned int timeo, unsigned int retrans)

{

	to->to_initval = timeo * HZ / 10;

/*

 * Create an RPC client handle

 */

static int nfs_create_rpc_client(struct nfs_client *clp,

int nfs_create_rpc_client(struct nfs_client *clp,

			  const struct rpc_timeout *timeparms,

			  rpc_authflavor_t flavor)

{

/*

 * Create a general RPC client

 */

static int nfs_init_server_rpcclient(struct nfs_server *server,

int nfs_init_server_rpcclient(struct nfs_server *server,

		const struct rpc_timeout *timeo,

		rpc_authflavor_t pseudoflavour)

{

		server->wsize = NFS_MAX_FILE_IO_SIZE;

	server->wpages = (server->wsize + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;

	server->pnfs_blksize = fsinfo->blksize;

	set_pnfs_layoutdriver(server, mntfh, fsinfo->layouttype);

	server->wtmult = nfs_block_bits(fsinfo->wtmult, NULL);

/*

 * Probe filesystem information, including the FSID on v2/v3

 */

static int nfs_probe_fsinfo(struct nfs_server *server, struct nfs_fh *mntfh, struct nfs_fattr *fattr)

int nfs_probe_fsinfo(struct nfs_server *server, struct nfs_fh *mntfh, struct nfs_fattr *fattr)

{

	struct nfs_fsinfo fsinfo;

	struct nfs_client *clp = server->nfs_client;

/*

 * Copy useful information when duplicating a server record

 */

static void nfs_server_copy_userdata(struct nfs_server *target, struct nfs_server *source)

void nfs_server_copy_userdata(struct nfs_server *target, struct nfs_server *source)

{

	target->flags = source->flags;

	target->rsize = source->rsize;

	target->options = source->options;

}

static void nfs_server_insert_lists(struct nfs_server *server)

void nfs_server_insert_lists(struct nfs_server *server)

{

	struct nfs_client *clp = server->nfs_client;

	struct nfs_net *nn = net_generic(clp->cl_net, nfs_net_id);

/*

 * Allocate and initialise a server record

 */

static struct nfs_server *nfs_alloc_server(void)

struct nfs_server *nfs_alloc_server(void)

{

	struct nfs_server *server;

	dprintk("--> nfs_free_server()\n");

	nfs_server_remove_lists(server);

	unset_pnfs_layoutdriver(server);

	if (server->destroy != NULL)

		server->destroy(server);

	return ERR_PTR(error);

}

#ifdef CONFIG_NFS_V4

/*

 * NFSv4.0 callback thread helper

 *

 * Find a client by callback identifier

 */

struct nfs_client *

nfs4_find_client_ident(struct net *net, int cb_ident)

{

	struct nfs_client *clp;

	struct nfs_net *nn = net_generic(net, nfs_net_id);

	spin_lock(&nn->nfs_client_lock);

	clp = idr_find(&nn->cb_ident_idr, cb_ident);

	if (clp)

		atomic_inc(&clp->cl_count);

	spin_unlock(&nn->nfs_client_lock);

	return clp;

}

#if defined(CONFIG_NFS_V4_1)

/*

 * NFSv4.1 callback thread helper

 * For CB_COMPOUND calls, find a client by IP address, protocol version,

 * minorversion, and sessionID

 *

 * Returns NULL if no such client

 */

struct nfs_client *

nfs4_find_client_sessionid(struct net *net, const struct sockaddr *addr,

			   struct nfs4_sessionid *sid)

{

	struct nfs_client *clp;

	struct nfs_net *nn = net_generic(net, nfs_net_id);

	spin_lock(&nn->nfs_client_lock);

	list_for_each_entry(clp, &nn->nfs_client_list, cl_share_link) {

		if (nfs4_cb_match_client(addr, clp, 1) == false)

			continue;

		if (!nfs4_has_session(clp))

			continue;

		/* Match sessionid*/

		if (memcmp(clp->cl_session->sess_id.data,

		    sid->data, NFS4_MAX_SESSIONID_LEN) != 0)

			continue;

		atomic_inc(&clp->cl_count);

		spin_unlock(&nn->nfs_client_lock);

		return clp;

	}

	spin_unlock(&nn->nfs_client_lock);

	return NULL;

}

#else /* CONFIG_NFS_V4_1 */

struct nfs_client *

nfs4_find_client_sessionid(struct net *net, const struct sockaddr *addr,

			   struct nfs4_sessionid *sid)

{

	return NULL;

}

#endif /* CONFIG_NFS_V4_1 */

/*

 * Initialize the NFS4 callback service

 */

static int nfs4_init_callback(struct nfs_client *clp)

{

	int error;

	if (clp->rpc_ops->version == 4) {

		struct rpc_xprt *xprt;

		xprt = rcu_dereference_raw(clp->cl_rpcclient->cl_xprt);

		if (nfs4_has_session(clp)) {

			error = xprt_setup_backchannel(xprt,

						NFS41_BC_MIN_CALLBACKS);

			if (error < 0)

				return error;

		}

		error = nfs_callback_up(clp->cl_mvops->minor_version, xprt);

		if (error < 0) {

			dprintk("%s: failed to start callback. Error = %d\n",

				__func__, error);

			return error;

		}

		__set_bit(NFS_CS_CALLBACK, &clp->cl_res_state);

	}

	return 0;

}

/*

 * Initialize the minor version specific parts of an NFS4 client record

 */

static int nfs4_init_client_minor_version(struct nfs_client *clp)

{

#if defined(CONFIG_NFS_V4_1)

	if (clp->cl_mvops->minor_version) {

		struct nfs4_session *session = NULL;

		/*

		 * Create the session and mark it expired.

		 * When a SEQUENCE operation encounters the expired session

		 * it will do session recovery to initialize it.

		 */

		session = nfs4_alloc_session(clp);

		if (!session)

			return -ENOMEM;

		clp->cl_session = session;

		/*

		 * The create session reply races with the server back

		 * channel probe. Mark the client NFS_CS_SESSION_INITING

		 * so that the client back channel can find the

		 * nfs_client struct

		 */

		nfs_mark_client_ready(clp, NFS_CS_SESSION_INITING);

	}

#endif /* CONFIG_NFS_V4_1 */

	return nfs4_init_callback(clp);

}

/**

 * nfs4_init_client - Initialise an NFS4 client record

 *

 * @clp: nfs_client to initialise

 * @timeparms: timeout parameters for underlying RPC transport

 * @ip_addr: callback IP address in presentation format

 * @authflavor: authentication flavor for underlying RPC transport

 *

 * Returns pointer to an NFS client, or an ERR_PTR value.

 */

struct nfs_client *nfs4_init_client(struct nfs_client *clp,

				    const struct rpc_timeout *timeparms,

				    const char *ip_addr,

				    rpc_authflavor_t authflavour)

{

	char buf[INET6_ADDRSTRLEN + 1];

	int error;

	if (clp->cl_cons_state == NFS_CS_READY) {

		/* the client is initialised already */

		dprintk("<-- nfs4_init_client() = 0 [already %p]\n", clp);

		return clp;

	}

	/* Check NFS protocol revision and initialize RPC op vector */

	clp->rpc_ops = &nfs_v4_clientops;

	__set_bit(NFS_CS_DISCRTRY, &clp->cl_flags);

	error = nfs_create_rpc_client(clp, timeparms, authflavour);

	if (error < 0)

		goto error;

	/* If no clientaddr= option was specified, find a usable cb address */

	if (ip_addr == NULL) {

		struct sockaddr_storage cb_addr;

		struct sockaddr *sap = (struct sockaddr *)&cb_addr;

		error = rpc_localaddr(clp->cl_rpcclient, sap, sizeof(cb_addr));

		if (error < 0)

			goto error;

		error = rpc_ntop(sap, buf, sizeof(buf));

		if (error < 0)

			goto error;

		ip_addr = (const char *)buf;

	}

	strlcpy(clp->cl_ipaddr, ip_addr, sizeof(clp->cl_ipaddr));

	error = nfs_idmap_new(clp);

	if (error < 0) {

		dprintk("%s: failed to create idmapper. Error = %d\n",

			__func__, error);

		goto error;

	}

	__set_bit(NFS_CS_IDMAP, &clp->cl_res_state);

	error = nfs4_init_client_minor_version(clp);

	if (error < 0)

		goto error;

	if (!nfs4_has_session(clp))

		nfs_mark_client_ready(clp, NFS_CS_READY);

	return clp;

error:

	nfs_mark_client_ready(clp, error);

	nfs_put_client(clp);

	dprintk("<-- nfs4_init_client() = xerror %d\n", error);

	return ERR_PTR(error);

}

/*

 * Set up an NFS4 client

 */

static int nfs4_set_client(struct nfs_server *server,

		const char *hostname,

		const struct sockaddr *addr,

		const size_t addrlen,

		const char *ip_addr,

		rpc_authflavor_t authflavour,

		int proto, const struct rpc_timeout *timeparms,

		u32 minorversion, struct net *net)

{

	struct nfs_client_initdata cl_init = {

		.hostname = hostname,

		.addr = addr,

		.addrlen = addrlen,

		.rpc_ops = &nfs_v4_clientops,

		.proto = proto,

		.minorversion = minorversion,

		.net = net,

	};

	struct nfs_client *clp;

	int error;

	dprintk("--> nfs4_set_client()\n");

	if (server->flags & NFS_MOUNT_NORESVPORT)

		set_bit(NFS_CS_NORESVPORT, &cl_init.init_flags);

	/* Allocate or find a client reference we can use */

	clp = nfs_get_client(&cl_init, timeparms, ip_addr, authflavour);

	if (IS_ERR(clp)) {

		error = PTR_ERR(clp);

		goto error;

	}

	/*

	 * Query for the lease time on clientid setup or renewal

	 *

	 * Note that this will be set on nfs_clients that were created

	 * only for the DS role and did not set this bit, but now will

	 * serve a dual role.

	 */

	set_bit(NFS_CS_CHECK_LEASE_TIME, &clp->cl_res_state);

	server->nfs_client = clp;

	dprintk("<-- nfs4_set_client() = 0 [new %p]\n", clp);

	return 0;

error:

	dprintk("<-- nfs4_set_client() = xerror %d\n", error);

	return error;

}

/*

 * Set up a pNFS Data Server client.

 *

 * Return any existing nfs_client that matches server address,port,version

 * and minorversion.

 *

 * For a new nfs_client, use a soft mount (default), a low retrans and a

 * low timeout interval so that if a connection is lost, we retry through

 * the MDS.

 */

struct nfs_client *nfs4_set_ds_client(struct nfs_client* mds_clp,

		const struct sockaddr *ds_addr, int ds_addrlen,

		int ds_proto, unsigned int ds_timeo, unsigned int ds_retrans)

{

	struct nfs_client_initdata cl_init = {

		.addr = ds_addr,

		.addrlen = ds_addrlen,

		.rpc_ops = &nfs_v4_clientops,

		.proto = ds_proto,

		.minorversion = mds_clp->cl_minorversion,

		.net = mds_clp->cl_net,

	};

	struct rpc_timeout ds_timeout;

	struct nfs_client *clp;

	/*

	 * Set an authflavor equual to the MDS value. Use the MDS nfs_client

	 * cl_ipaddr so as to use the same EXCHANGE_ID co_ownerid as the MDS

	 * (section 13.1 RFC 5661).

	 */

	nfs_init_timeout_values(&ds_timeout, ds_proto, ds_timeo, ds_retrans);

	clp = nfs_get_client(&cl_init, &ds_timeout, mds_clp->cl_ipaddr,

			     mds_clp->cl_rpcclient->cl_auth->au_flavor);

	dprintk("<-- %s %p\n", __func__, clp);

	return clp;

}

EXPORT_SYMBOL_GPL(nfs4_set_ds_client);

/*

 * Session has been established, and the client marked ready.

 * Set the mount rsize and wsize with negotiated fore channel

 * attributes which will be bound checked in nfs_server_set_fsinfo.

 */

static void nfs4_session_set_rwsize(struct nfs_server *server)

{

#ifdef CONFIG_NFS_V4_1

	struct nfs4_session *sess;

	u32 server_resp_sz;

	u32 server_rqst_sz;

	if (!nfs4_has_session(server->nfs_client))

		return;

	sess = server->nfs_client->cl_session;

	server_resp_sz = sess->fc_attrs.max_resp_sz - nfs41_maxread_overhead;

	server_rqst_sz = sess->fc_attrs.max_rqst_sz - nfs41_maxwrite_overhead;

	if (server->rsize > server_resp_sz)

		server->rsize = server_resp_sz;

	if (server->wsize > server_rqst_sz)

		server->wsize = server_rqst_sz;

#endif /* CONFIG_NFS_V4_1 */

}

static int nfs4_server_common_setup(struct nfs_server *server,

		struct nfs_fh *mntfh)

{

	struct nfs_fattr *fattr;

	int error;

	BUG_ON(!server->nfs_client);

	BUG_ON(!server->nfs_client->rpc_ops);

	BUG_ON(!server->nfs_client->rpc_ops->file_inode_ops);

	/* data servers support only a subset of NFSv4.1 */

	if (is_ds_only_client(server->nfs_client))

		return -EPROTONOSUPPORT;