diff options
Diffstat (limited to 'fs/nfsd/nfs4state.c')
-rw-r--r-- | fs/nfsd/nfs4state.c | 924 |
1 files changed, 568 insertions, 356 deletions
diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 4a2734758778..f1e5ec6b5105 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -33,7 +33,7 @@ */ #include <linux/file.h> -#include <linux/smp_lock.h> +#include <linux/fs.h> #include <linux/slab.h> #include <linux/namei.h> #include <linux/swap.h> @@ -51,7 +51,6 @@ static time_t boot_time; static u32 current_ownerid = 1; static u32 current_fileid = 1; static u32 current_delegid = 1; -static u32 nfs4_init; static stateid_t zerostateid; /* bits all 0 */ static stateid_t onestateid; /* bits all 1 */ static u64 current_sessionid = 1; @@ -163,14 +162,60 @@ static struct list_head ownerstr_hashtbl[OWNER_HASH_SIZE]; static struct list_head file_hashtbl[FILE_HASH_SIZE]; static struct list_head stateid_hashtbl[STATEID_HASH_SIZE]; +static void __nfs4_file_get_access(struct nfs4_file *fp, int oflag) +{ + BUG_ON(!(fp->fi_fds[oflag] || fp->fi_fds[O_RDWR])); + atomic_inc(&fp->fi_access[oflag]); +} + +static void nfs4_file_get_access(struct nfs4_file *fp, int oflag) +{ + if (oflag == O_RDWR) { + __nfs4_file_get_access(fp, O_RDONLY); + __nfs4_file_get_access(fp, O_WRONLY); + } else + __nfs4_file_get_access(fp, oflag); +} + +static void nfs4_file_put_fd(struct nfs4_file *fp, int oflag) +{ + if (fp->fi_fds[oflag]) { + fput(fp->fi_fds[oflag]); + fp->fi_fds[oflag] = NULL; + } +} + +static void __nfs4_file_put_access(struct nfs4_file *fp, int oflag) +{ + if (atomic_dec_and_test(&fp->fi_access[oflag])) { + nfs4_file_put_fd(fp, O_RDWR); + nfs4_file_put_fd(fp, oflag); + } +} + +static void nfs4_file_put_access(struct nfs4_file *fp, int oflag) +{ + if (oflag == O_RDWR) { + __nfs4_file_put_access(fp, O_RDONLY); + __nfs4_file_put_access(fp, O_WRONLY); + } else + __nfs4_file_put_access(fp, oflag); +} + static struct nfs4_delegation * alloc_init_deleg(struct nfs4_client *clp, struct nfs4_stateid *stp, struct svc_fh *current_fh, u32 type) { struct nfs4_delegation *dp; struct nfs4_file *fp = stp->st_file; - struct nfs4_cb_conn *cb = &stp->st_stateowner->so_client->cl_cb_conn; dprintk("NFSD alloc_init_deleg\n"); + /* + * Major work on the lease subsystem (for example, to support + * calbacks on stat) will be required before we can support + * write delegations properly. + */ + if (type != NFS4_OPEN_DELEGATE_READ) + return NULL; if (fp->fi_had_conflict) return NULL; if (num_delegations > max_delegations) @@ -185,11 +230,9 @@ alloc_init_deleg(struct nfs4_client *clp, struct nfs4_stateid *stp, struct svc_f dp->dl_client = clp; get_nfs4_file(fp); dp->dl_file = fp; + nfs4_file_get_access(fp, O_RDONLY); dp->dl_flock = NULL; - get_file(stp->st_vfs_file); - dp->dl_vfs_file = stp->st_vfs_file; dp->dl_type = type; - dp->dl_ident = cb->cb_ident; dp->dl_stateid.si_boot = boot_time; dp->dl_stateid.si_stateownerid = current_delegid++; dp->dl_stateid.si_fileid = 0; @@ -222,15 +265,12 @@ nfs4_put_delegation(struct nfs4_delegation *dp) static void nfs4_close_delegation(struct nfs4_delegation *dp) { - struct file *filp = dp->dl_vfs_file; + struct file *filp = find_readable_file(dp->dl_file); dprintk("NFSD: close_delegation dp %p\n",dp); - dp->dl_vfs_file = NULL; - /* The following nfsd_close may not actually close the file, - * but we want to remove the lease in any case. */ if (dp->dl_flock) vfs_setlease(filp, F_UNLCK, &dp->dl_flock); - nfsd_close(filp); + nfs4_file_put_access(dp->dl_file, O_RDONLY); } /* Called under the state lock. */ @@ -302,8 +342,12 @@ static void free_generic_stateid(struct nfs4_stateid *stp) static void release_lock_stateid(struct nfs4_stateid *stp) { + struct file *file; + unhash_generic_stateid(stp); - locks_remove_posix(stp->st_vfs_file, (fl_owner_t)stp->st_stateowner); + file = find_any_file(stp->st_file); + if (file) + locks_remove_posix(file, (fl_owner_t)stp->st_stateowner); free_generic_stateid(stp); } @@ -341,11 +385,85 @@ release_stateid_lockowners(struct nfs4_stateid *open_stp) } } +/* + * We store the NONE, READ, WRITE, and BOTH bits separately in the + * st_{access,deny}_bmap field of the stateid, in order to track not + * only what share bits are currently in force, but also what + * combinations of share bits previous opens have used. This allows us + * to enforce the recommendation of rfc 3530 14.2.19 that the server + * return an error if the client attempt to downgrade to a combination + * of share bits not explicable by closing some of its previous opens. + * + * XXX: This enforcement is actually incomplete, since we don't keep + * track of access/deny bit combinations; so, e.g., we allow: + * + * OPEN allow read, deny write + * OPEN allow both, deny none + * DOWNGRADE allow read, deny none + * + * which we should reject. + */ +static void +set_access(unsigned int *access, unsigned long bmap) { + int i; + + *access = 0; + for (i = 1; i < 4; i++) { + if (test_bit(i, &bmap)) + *access |= i; + } +} + +static void +set_deny(unsigned int *deny, unsigned long bmap) { + int i; + + *deny = 0; + for (i = 0; i < 4; i++) { + if (test_bit(i, &bmap)) + *deny |= i ; + } +} + +static int +test_share(struct nfs4_stateid *stp, struct nfsd4_open *open) { + unsigned int access, deny; + + set_access(&access, stp->st_access_bmap); + set_deny(&deny, stp->st_deny_bmap); + if ((access & open->op_share_deny) || (deny & open->op_share_access)) + return 0; + return 1; +} + +static int nfs4_access_to_omode(u32 access) +{ + switch (access & NFS4_SHARE_ACCESS_BOTH) { + case NFS4_SHARE_ACCESS_READ: + return O_RDONLY; + case NFS4_SHARE_ACCESS_WRITE: + return O_WRONLY; + case NFS4_SHARE_ACCESS_BOTH: + return O_RDWR; + } + BUG(); +} + +static int nfs4_access_bmap_to_omode(struct nfs4_stateid *stp) +{ + unsigned int access; + + set_access(&access, stp->st_access_bmap); + return nfs4_access_to_omode(access); +} + static void release_open_stateid(struct nfs4_stateid *stp) { + int oflag = nfs4_access_bmap_to_omode(stp); + unhash_generic_stateid(stp); release_stateid_lockowners(stp); - nfsd_close(stp->st_vfs_file); + nfs4_file_put_access(stp->st_file, oflag); free_generic_stateid(stp); } @@ -415,171 +533,258 @@ gen_sessionid(struct nfsd4_session *ses) */ #define NFSD_MIN_HDR_SEQ_SZ (24 + 12 + 44) +static void +free_session_slots(struct nfsd4_session *ses) +{ + int i; + + for (i = 0; i < ses->se_fchannel.maxreqs; i++) + kfree(ses->se_slots[i]); +} + /* - * Give the client the number of ca_maxresponsesize_cached slots it - * requests, of size bounded by NFSD_SLOT_CACHE_SIZE, - * NFSD_MAX_MEM_PER_SESSION, and nfsd_drc_max_mem. Do not allow more - * than NFSD_MAX_SLOTS_PER_SESSION. - * - * If we run out of reserved DRC memory we should (up to a point) + * We don't actually need to cache the rpc and session headers, so we + * can allocate a little less for each slot: + */ +static inline int slot_bytes(struct nfsd4_channel_attrs *ca) +{ + return ca->maxresp_cached - NFSD_MIN_HDR_SEQ_SZ; +} + +static int nfsd4_sanitize_slot_size(u32 size) +{ + size -= NFSD_MIN_HDR_SEQ_SZ; /* We don't cache the rpc header */ + size = min_t(u32, size, NFSD_SLOT_CACHE_SIZE); + + return size; +} + +/* + * XXX: If we run out of reserved DRC memory we could (up to a point) * re-negotiate active sessions and reduce their slot usage to make * rooom for new connections. For now we just fail the create session. */ -static int set_forechannel_drc_size(struct nfsd4_channel_attrs *fchan) +static int nfsd4_get_drc_mem(int slotsize, u32 num) { - int mem, size = fchan->maxresp_cached; + int avail; - if (fchan->maxreqs < 1) - return nfserr_inval; + num = min_t(u32, num, NFSD_MAX_SLOTS_PER_SESSION); - if (size < NFSD_MIN_HDR_SEQ_SZ) - size = NFSD_MIN_HDR_SEQ_SZ; - size -= NFSD_MIN_HDR_SEQ_SZ; - if (size > NFSD_SLOT_CACHE_SIZE) - size = NFSD_SLOT_CACHE_SIZE; - - /* bound the maxreqs by NFSD_MAX_MEM_PER_SESSION */ - mem = fchan->maxreqs * size; - if (mem > NFSD_MAX_MEM_PER_SESSION) { - fchan->maxreqs = NFSD_MAX_MEM_PER_SESSION / size; - if (fchan->maxreqs > NFSD_MAX_SLOTS_PER_SESSION) - fchan->maxreqs = NFSD_MAX_SLOTS_PER_SESSION; - mem = fchan->maxreqs * size; - } + spin_lock(&nfsd_drc_lock); + avail = min_t(int, NFSD_MAX_MEM_PER_SESSION, + nfsd_drc_max_mem - nfsd_drc_mem_used); + num = min_t(int, num, avail / slotsize); + nfsd_drc_mem_used += num * slotsize; + spin_unlock(&nfsd_drc_lock); + + return num; +} +static void nfsd4_put_drc_mem(int slotsize, int num) +{ spin_lock(&nfsd_drc_lock); - /* bound the total session drc memory ussage */ - if (mem + nfsd_drc_mem_used > nfsd_drc_max_mem) { - fchan->maxreqs = (nfsd_drc_max_mem - nfsd_drc_mem_used) / size; - mem = fchan->maxreqs * size; - } - nfsd_drc_mem_used += mem; + nfsd_drc_mem_used -= slotsize * num; spin_unlock(&nfsd_drc_lock); +} - if (fchan->maxreqs == 0) - return nfserr_serverfault; +static struct nfsd4_session *alloc_session(int slotsize, int numslots) +{ + struct nfsd4_session *new; + int mem, i; - fchan->maxresp_cached = size + NFSD_MIN_HDR_SEQ_SZ; - return 0; + BUILD_BUG_ON(NFSD_MAX_SLOTS_PER_SESSION * sizeof(struct nfsd4_slot *) + + sizeof(struct nfsd4_session) > PAGE_SIZE); + mem = numslots * sizeof(struct nfsd4_slot *); + + new = kzalloc(sizeof(*new) + mem, GFP_KERNEL); + if (!new) + return NULL; + /* allocate each struct nfsd4_slot and data cache in one piece */ + for (i = 0; i < numslots; i++) { + mem = sizeof(struct nfsd4_slot) + slotsize; + new->se_slots[i] = kzalloc(mem, GFP_KERNEL); + if (!new->se_slots[i]) + goto out_free; + } + return new; +out_free: + while (i--) + kfree(new->se_slots[i]); + kfree(new); + return NULL; } -/* - * fchan holds the client values on input, and the server values on output - * sv_max_mesg is the maximum payload plus one page for overhead. - */ -static int init_forechannel_attrs(struct svc_rqst *rqstp, - struct nfsd4_channel_attrs *session_fchan, - struct nfsd4_channel_attrs *fchan) +static void init_forechannel_attrs(struct nfsd4_channel_attrs *new, struct nfsd4_channel_attrs *req, int numslots, int slotsize) { - int status = 0; - __u32 maxcount = nfsd_serv->sv_max_mesg; + u32 maxrpc = nfsd_serv->sv_max_mesg; - /* headerpadsz set to zero in encode routine */ + new->maxreqs = numslots; + new->maxresp_cached = slotsize + NFSD_MIN_HDR_SEQ_SZ; + new->maxreq_sz = min_t(u32, req->maxreq_sz, maxrpc); + new->maxresp_sz = min_t(u32, req->maxresp_sz, maxrpc); + new->maxops = min_t(u32, req->maxops, NFSD_MAX_OPS_PER_COMPOUND); +} - /* Use the client's max request and max response size if possible */ - if (fchan->maxreq_sz > maxcount) - fchan->maxreq_sz = maxcount; - session_fchan->maxreq_sz = fchan->maxreq_sz; +static void free_conn(struct nfsd4_conn *c) +{ + svc_xprt_put(c->cn_xprt); + kfree(c); +} - if (fchan->maxresp_sz > maxcount) - fchan->maxresp_sz = maxcount; - session_fchan->maxresp_sz = fchan->maxresp_sz; +static void nfsd4_conn_lost(struct svc_xpt_user *u) +{ + struct nfsd4_conn *c = container_of(u, struct nfsd4_conn, cn_xpt_user); + struct nfs4_client *clp = c->cn_session->se_client; - /* Use the client's maxops if possible */ - if (fchan->maxops > NFSD_MAX_OPS_PER_COMPOUND) - fchan->maxops = NFSD_MAX_OPS_PER_COMPOUND; - session_fchan->maxops = fchan->maxops; + spin_lock(&clp->cl_lock); + if (!list_empty(&c->cn_persession)) { + list_del(&c->cn_persession); + free_conn(c); + } + spin_unlock(&clp->cl_lock); +} - /* FIXME: Error means no more DRC pages so the server should - * recover pages from existing sessions. For now fail session - * creation. - */ - status = set_forechannel_drc_size(fchan); +static struct nfsd4_conn *alloc_conn(struct svc_rqst *rqstp, u32 flags) +{ + struct nfsd4_conn *conn; - session_fchan->maxresp_cached = fchan->maxresp_cached; - session_fchan->maxreqs = fchan->maxreqs; + conn = kmalloc(sizeof(struct nfsd4_conn), GFP_KERNEL); + if (!conn) + return NULL; + svc_xprt_get(rqstp->rq_xprt); + conn->cn_xprt = rqstp->rq_xprt; + conn->cn_flags = flags; + INIT_LIST_HEAD(&conn->cn_xpt_user.list); + return conn; +} - dprintk("%s status %d\n", __func__, status); - return status; +static void __nfsd4_hash_conn(struct nfsd4_conn *conn, struct nfsd4_session *ses) +{ + conn->cn_session = ses; + list_add(&conn->cn_persession, &ses->se_conns); } -static void -free_session_slots(struct nfsd4_session *ses) +static void nfsd4_hash_conn(struct nfsd4_conn *conn, struct nfsd4_session *ses) { - int i; + struct nfs4_client *clp = ses->se_client; - for (i = 0; i < ses->se_fchannel.maxreqs; i++) - kfree(ses->se_slots[i]); + spin_lock(&clp->cl_lock); + __nfsd4_hash_conn(conn, ses); + spin_unlock(&clp->cl_lock); } -/* - * We don't actually need to cache the rpc and session headers, so we - * can allocate a little less for each slot: - */ -static inline int slot_bytes(struct nfsd4_channel_attrs *ca) +static void nfsd4_register_conn(struct nfsd4_conn *conn) { - return ca->maxresp_cached - NFSD_MIN_HDR_SEQ_SZ; + conn->cn_xpt_user.callback = nfsd4_conn_lost; + register_xpt_user(conn->cn_xprt, &conn->cn_xpt_user); } -static int -alloc_init_session(struct svc_rqst *rqstp, struct nfs4_client *clp, - struct nfsd4_create_session *cses) +static __be32 nfsd4_new_conn(struct svc_rqst *rqstp, struct nfsd4_session *ses) { - struct nfsd4_session *new, tmp; - struct nfsd4_slot *sp; - int idx, slotsize, cachesize, i; - int status; + struct nfsd4_conn *conn; + u32 flags = NFS4_CDFC4_FORE; - memset(&tmp, 0, sizeof(tmp)); + if (ses->se_flags & SESSION4_BACK_CHAN) + flags |= NFS4_CDFC4_BACK; + conn = alloc_conn(rqstp, flags); + if (!conn) + return nfserr_jukebox; + nfsd4_hash_conn(conn, ses); + nfsd4_register_conn(conn); + return nfs_ok; +} - /* FIXME: For now, we just accept the client back channel attributes. */ - tmp.se_bchannel = cses->back_channel; - status = init_forechannel_attrs(rqstp, &tmp.se_fchannel, - &cses->fore_channel); - if (status) - goto out; +static void nfsd4_del_conns(struct nfsd4_session *s) +{ + struct nfs4_client *clp = s->se_client; + struct nfsd4_conn *c; - BUILD_BUG_ON(NFSD_MAX_SLOTS_PER_SESSION * sizeof(struct nfsd4_slot) - + sizeof(struct nfsd4_session) > PAGE_SIZE); + spin_lock(&clp->cl_lock); + while (!list_empty(&s->se_conns)) { + c = list_first_entry(&s->se_conns, struct nfsd4_conn, cn_persession); + list_del_init(&c->cn_persession); + spin_unlock(&clp->cl_lock); - status = nfserr_serverfault; - /* allocate struct nfsd4_session and slot table pointers in one piece */ - slotsize = tmp.se_fchannel.maxreqs * sizeof(struct nfsd4_slot *); - new = kzalloc(sizeof(*new) + slotsize, GFP_KERNEL); - if (!new) - goto out; + unregister_xpt_user(c->cn_xprt, &c->cn_xpt_user); + free_conn(c); - memcpy(new, &tmp, sizeof(*new)); + spin_lock(&clp->cl_lock); + } + spin_unlock(&clp->cl_lock); +} - /* allocate each struct nfsd4_slot and data cache in one piece */ - cachesize = slot_bytes(&new->se_fchannel); - for (i = 0; i < new->se_fchannel.maxreqs; i++) { - sp = kzalloc(sizeof(*sp) + cachesize, GFP_KERNEL); - if (!sp) - goto out_free; - new->se_slots[i] = sp; +void free_session(struct kref *kref) +{ + struct nfsd4_session *ses; + int mem; + + ses = container_of(kref, struct nfsd4_session, se_ref); + nfsd4_del_conns(ses); + spin_lock(&nfsd_drc_lock); + mem = ses->se_fchannel.maxreqs * slot_bytes(&ses->se_fchannel); + nfsd_drc_mem_used -= mem; + spin_unlock(&nfsd_drc_lock); + free_session_slots(ses); + kfree(ses); +} + +static struct nfsd4_session *alloc_init_session(struct svc_rqst *rqstp, struct nfs4_client *clp, struct nfsd4_create_session *cses) +{ + struct nfsd4_session *new; + struct nfsd4_channel_attrs *fchan = &cses->fore_channel; + int numslots, slotsize; + int status; + int idx; + + /* + * Note decreasing slot size below client's request may + * make it difficult for client to function correctly, whereas + * decreasing the number of slots will (just?) affect + * performance. When short on memory we therefore prefer to + * decrease number of slots instead of their size. + */ + slotsize = nfsd4_sanitize_slot_size(fchan->maxresp_cached); + numslots = nfsd4_get_drc_mem(slotsize, fchan->maxreqs); + + new = alloc_session(slotsize, numslots); + if (!new) { + nfsd4_put_drc_mem(slotsize, fchan->maxreqs); + return NULL; } + init_forechannel_attrs(&new->se_fchannel, fchan, numslots, slotsize); new->se_client = clp; gen_sessionid(new); - idx = hash_sessionid(&new->se_sessionid); - memcpy(clp->cl_sessionid.data, new->se_sessionid.data, - NFS4_MAX_SESSIONID_LEN); + INIT_LIST_HEAD(&new->se_conns); + + new->se_cb_seq_nr = 1; new->se_flags = cses->flags; + new->se_cb_prog = cses->callback_prog; kref_init(&new->se_ref); + idx = hash_sessionid(&new->se_sessionid); spin_lock(&client_lock); list_add(&new->se_hash, &sessionid_hashtbl[idx]); list_add(&new->se_perclnt, &clp->cl_sessions); spin_unlock(&client_lock); - status = nfs_ok; -out: - return status; -out_free: - free_session_slots(new); - kfree(new); - goto out; + status = nfsd4_new_conn(rqstp, new); + /* whoops: benny points out, status is ignored! (err, or bogus) */ + if (status) { + free_session(&new->se_ref); + return NULL; + } + if (!clp->cl_cb_session && (cses->flags & SESSION4_BACK_CHAN)) { + struct sockaddr *sa = svc_addr(rqstp); + + clp->cl_cb_session = new; + clp->cl_cb_conn.cb_xprt = rqstp->rq_xprt; + svc_xprt_get(rqstp->rq_xprt); + rpc_copy_addr((struct sockaddr *)&clp->cl_cb_conn.cb_addr, sa); + clp->cl_cb_conn.cb_addrlen = svc_addr_len(sa); + nfsd4_probe_callback(clp); + } + return new; } /* caller must hold client_lock */ @@ -591,10 +796,8 @@ find_in_sessionid_hashtbl(struct nfs4_sessionid *sessionid) dump_sessionid(__func__, sessionid); idx = hash_sessionid(sessionid); - dprintk("%s: idx is %d\n", __func__, idx); /* Search in the appropriate list */ list_for_each_entry(elem, &sessionid_hashtbl[idx], se_hash) { - dump_sessionid("list traversal", &elem->se_sessionid); if (!memcmp(elem->se_sessionid.data, sessionid->data, NFS4_MAX_SESSIONID_LEN)) { return elem; @@ -613,21 +816,6 @@ unhash_session(struct nfsd4_session *ses) list_del(&ses->se_perclnt); } -void -free_session(struct kref *kref) -{ - struct nfsd4_session *ses; - int mem; - - ses = container_of(kref, struct nfsd4_session, se_ref); - spin_lock(&nfsd_drc_lock); - mem = ses->se_fchannel.maxreqs * slot_bytes(&ses->se_fchannel); - nfsd_drc_mem_used -= mem; - spin_unlock(&nfsd_drc_lock); - free_session_slots(ses); - kfree(ses); -} - /* must be called under the client_lock */ static inline void renew_client_locked(struct nfs4_client *clp) @@ -694,6 +882,13 @@ static struct nfs4_client *alloc_client(struct xdr_netobj name) static inline void free_client(struct nfs4_client *clp) { + while (!list_empty(&clp->cl_sessions)) { + struct nfsd4_session *ses; + ses = list_entry(clp->cl_sessions.next, struct nfsd4_session, + se_perclnt); + list_del(&ses->se_perclnt); + nfsd4_put_session(ses); + } if (clp->cl_cred.cr_group_info) put_group_info(clp->cl_cred.cr_group_info); kfree(clp->cl_principal); @@ -714,22 +909,18 @@ release_session_client(struct nfsd4_session *session) } else renew_client_locked(clp); spin_unlock(&client_lock); - nfsd4_put_session(session); } /* must be called under the client_lock */ static inline void unhash_client_locked(struct nfs4_client *clp) { + struct nfsd4_session *ses; + mark_client_expired(clp); list_del(&clp->cl_lru); - while (!list_empty(&clp->cl_sessions)) { - struct nfsd4_session *ses; - ses = list_entry(clp->cl_sessions.next, struct nfsd4_session, - se_perclnt); - unhash_session(ses); - nfsd4_put_session(ses); - } + list_for_each_entry(ses, &clp->cl_sessions, se_perclnt) + list_del_init(&ses->se_hash); } static void @@ -758,7 +949,7 @@ expire_client(struct nfs4_client *clp) sop = list_entry(clp->cl_openowners.next, struct nfs4_stateowner, so_perclient); release_openowner(sop); } - nfsd4_set_callback_client(clp, NULL); + nfsd4_shutdown_callback(clp); if (clp->cl_cb_conn.cb_xprt) svc_xprt_put(clp->cl_cb_conn.cb_xprt); list_del(&clp->cl_idhash); @@ -843,6 +1034,8 @@ static struct nfs4_client *create_client(struct xdr_netobj name, char *recdir, if (clp == NULL) return NULL; + INIT_LIST_HEAD(&clp->cl_sessions); + princ = svc_gss_principal(rqstp); if (princ) { clp->cl_principal = kstrdup(princ, GFP_KERNEL); @@ -859,8 +1052,9 @@ static struct nfs4_client *create_client(struct xdr_netobj name, char *recdir, INIT_LIST_HEAD(&clp->cl_strhash); INIT_LIST_HEAD(&clp->cl_openowners); INIT_LIST_HEAD(&clp->cl_delegations); - INIT_LIST_HEAD(&clp->cl_sessions); INIT_LIST_HEAD(&clp->cl_lru); + spin_lock_init(&clp->cl_lock); + INIT_WORK(&clp->cl_cb_null.cb_work, nfsd4_do_callback_rpc); clp->cl_time = get_seconds(); clear_bit(0, &clp->cl_cb_slot_busy); rpc_init_wait_queue(&clp->cl_cb_waitq, "Backchannel slot table"); @@ -869,7 +1063,7 @@ static struct nfs4_client *create_client(struct xdr_netobj name, char *recdir, clp->cl_flavor = rqstp->rq_flavor; copy_cred(&clp->cl_cred, &rqstp->rq_cred); gen_confirm(clp); - + clp->cl_cb_session = NULL; return clp; } @@ -981,7 +1175,7 @@ find_unconfirmed_client_by_str(const char *dname, unsigned int hashval, static void gen_callback(struct nfs4_client *clp, struct nfsd4_setclientid *se, u32 scopeid) { - struct nfs4_cb_conn *cb = &clp->cl_cb_conn; + struct nfs4_cb_conn *conn = &clp->cl_cb_conn; unsigned short expected_family; /* Currently, we only support tcp and tcp6 for the callback channel */ @@ -994,24 +1188,23 @@ gen_callback(struct nfs4_client *clp, struct nfsd4_setclientid *se, u32 scopeid) else goto out_err; - cb->cb_addrlen = rpc_uaddr2sockaddr(se->se_callback_addr_val, + conn->cb_addrlen = rpc_uaddr2sockaddr(se->se_callback_addr_val, se->se_callback_addr_len, - (struct sockaddr *) &cb->cb_addr, - sizeof(cb->cb_addr)); + (struct sockaddr *)&conn->cb_addr, + sizeof(conn->cb_addr)); - if (!cb->cb_addrlen || cb->cb_addr.ss_family != expected_family) + if (!conn->cb_addrlen || conn->cb_addr.ss_family != expected_family) goto out_err; - if (cb->cb_addr.ss_family == AF_INET6) - ((struct sockaddr_in6 *) &cb->cb_addr)->sin6_scope_id = scopeid; + if (conn->cb_addr.ss_family == AF_INET6) + ((struct sockaddr_in6 *)&conn->cb_addr)->sin6_scope_id = scopeid; - cb->cb_minorversion = 0; - cb->cb_prog = se->se_callback_prog; - cb->cb_ident = se->se_callback_ident; + conn->cb_prog = se->se_callback_prog; + conn->cb_ident = se->se_callback_ident; return; out_err: - cb->cb_addr.ss_family = AF_UNSPEC; - cb->cb_addrlen = 0; + conn->cb_addr.ss_family = AF_UNSPEC; + conn->cb_addrlen = 0; dprintk(KERN_INFO "NFSD: this client (clientid %08x/%08x) " "will not receive delegations\n", clp->cl_clientid.cl_boot, clp->cl_clientid.cl_id); @@ -1220,7 +1413,7 @@ out_new: /* Normal case */ new = create_client(exid->clname, dname, rqstp, &verf); if (new == NULL) { - status = nfserr_serverfault; + status = nfserr_jukebox; goto out; } @@ -1298,7 +1491,9 @@ nfsd4_create_session(struct svc_rqst *rqstp, { struct sockaddr *sa = svc_addr(rqstp); struct nfs4_client *conf, *unconf; + struct nfsd4_session *new; struct nfsd4_clid_slot *cs_slot = NULL; + bool confirm_me = false; int status = 0; nfs4_lock_state(); @@ -1321,7 +1516,6 @@ nfsd4_create_session(struct svc_rqst *rqstp, cs_slot->sl_seqid, cr_ses->seqid); goto out; } - cs_slot->sl_seqid++; } else if (unconf) { if (!same_creds(&unconf->cl_cred, &rqstp->rq_cred) || !rpc_cmp_addr(sa, (struct sockaddr *) &unconf->cl_addr)) { @@ -1334,25 +1528,10 @@ nfsd4_create_session(struct svc_rqst *rqstp, if (status) { /* an unconfirmed replay returns misordered */ status = nfserr_seq_misordered; - goto out_cache; + goto out; } - cs_slot->sl_seqid++; /* from 0 to 1 */ - move_to_confirmed(unconf); - - if (cr_ses->flags & SESSION4_BACK_CHAN) { - unconf->cl_cb_conn.cb_xprt = rqstp->rq_xprt; - svc_xprt_get(rqstp->rq_xprt); - rpc_copy_addr( - (struct sockaddr *)&unconf->cl_cb_conn.cb_addr, - sa); - unconf->cl_cb_conn.cb_addrlen = svc_addr_len(sa); - unconf->cl_cb_conn.cb_minorversion = - cstate->minorversion; - unconf->cl_cb_conn.cb_prog = cr_ses->callback_prog; - unconf->cl_cb_seq_nr = 1; - nfsd4_probe_callback(unconf, &unconf->cl_cb_conn); - } + confirm_me = true; conf = unconf; } else { status = nfserr_stale_clientid; @@ -1360,22 +1539,30 @@ nfsd4_create_session(struct svc_rqst *rqstp, } /* + * XXX: we should probably set this at creation time, and check + * for consistent minorversion use throughout: + */ + conf->cl_minorversion = 1; + /* * We do not support RDMA or persistent sessions */ cr_ses->flags &= ~SESSION4_PERSIST; cr_ses->flags &= ~SESSION4_RDMA; - status = alloc_init_session(rqstp, conf, cr_ses); - if (status) + status = nfserr_jukebox; + new = alloc_init_session(rqstp, conf, cr_ses); + if (!new) goto out; - - memcpy(cr_ses->sessionid.data, conf->cl_sessionid.data, + status = nfs_ok; + memcpy(cr_ses->sessionid.data, new->se_sessionid.data, NFS4_MAX_SESSIONID_LEN); + cs_slot->sl_seqid++; cr_ses->seqid = cs_slot->sl_seqid; -out_cache: /* cache solo and embedded create sessions under the state lock */ nfsd4_cache_create_session(cr_ses, cs_slot, status); + if (confirm_me) + move_to_confirmed(conf); out: nfs4_unlock_state(); dprintk("%s returns %d\n", __func__, ntohl(status)); @@ -1429,8 +1616,11 @@ nfsd4_destroy_session(struct svc_rqst *r, nfs4_lock_state(); /* wait for callbacks */ - nfsd4_set_callback_client(ses->se_client, NULL); + nfsd4_shutdown_callback(ses->se_client); nfs4_unlock_state(); + + nfsd4_del_conns(ses); + nfsd4_put_session(ses); status = nfs_ok; out: @@ -1438,6 +1628,36 @@ out: return status; } +static struct nfsd4_conn *__nfsd4_find_conn(struct svc_xprt *xpt, struct nfsd4_session *s) +{ + struct nfsd4_conn *c; + + list_for_each_entry(c, &s->se_conns, cn_persession) { + if (c->cn_xprt == xpt) { + return c; + } + } + return NULL; +} + +static void nfsd4_sequence_check_conn(struct nfsd4_conn *new, struct nfsd4_session *ses) +{ + struct nfs4_client *clp = ses->se_client; + struct nfsd4_conn *c; + + spin_lock(&clp->cl_lock); + c = __nfsd4_find_conn(new->cn_xprt, ses); + if (c) { + spin_unlock(&clp->cl_lock); + free_conn(new); + return; + } + __nfsd4_hash_conn(new, ses); + spin_unlock(&clp->cl_lock); + nfsd4_register_conn(new); + return; +} + __be32 nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, @@ -1446,11 +1666,20 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compoundres *resp = rqstp->rq_resp; struct nfsd4_session *session; struct nfsd4_slot *slot; + struct nfsd4_conn *conn; int status; if (resp->opcnt != 1) return nfserr_sequence_pos; + /* + * Will be either used or freed by nfsd4_sequence_check_conn + * below. + */ + conn = alloc_conn(rqstp, NFS4_CDFC4_FORE); + if (!conn) + return nfserr_jukebox; + spin_lock(&client_lock); status = nfserr_badsession; session = find_in_sessionid_hashtbl(&seq->sessionid); @@ -1482,6 +1711,9 @@ nfsd4_sequence(struct svc_rqst *rqstp, if (status) goto out; + nfsd4_sequence_check_conn(conn, session); + conn = NULL; + /* Success! bump slot seqid */ slot->sl_inuse = true; slot->sl_seqid = seq->seqid; @@ -1496,6 +1728,7 @@ out: nfsd4_get_session(cstate->session); atomic_inc(&session->se_client->cl_refcount); } + kfree(conn); spin_unlock(&client_lock); dprintk("%s: return %d\n", __func__, ntohl(status)); return status; @@ -1630,6 +1863,11 @@ nfsd4_setclientid(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, goto out; gen_clid(new); } + /* + * XXX: we should probably set this at creation time, and check + * for consistent minorversion use throughout: + */ + new->cl_minorversion = 0; gen_callback(new, setclid, rpc_get_scope_id(sa)); add_to_unconfirmed(new, strhashval); setclid->se_clientid.cl_boot = new->cl_clientid.cl_boot; @@ -1690,7 +1928,8 @@ nfsd4_setclientid_confirm(struct svc_rqst *rqstp, status = nfserr_clid_inuse; else { atomic_set(&conf->cl_cb_set, 0); - nfsd4_probe_callback(conf, &unconf->cl_cb_conn); + nfsd4_change_callback(conf, &unconf->cl_cb_conn); + nfsd4_probe_callback(conf); expire_client(unconf); status = nfs_ok; @@ -1724,7 +1963,7 @@ nfsd4_setclientid_confirm(struct svc_rqst *rqstp, } move_to_confirmed(unconf); conf = unconf; - nfsd4_probe_callback(conf, &conf->cl_cb_conn); + nfsd4_probe_callback(conf); status = nfs_ok; } } else if ((!conf || (conf && !same_verf(&conf->cl_confirm, &confirm))) @@ -1760,6 +1999,8 @@ alloc_init_file(struct inode *ino) fp->fi_inode = igrab(ino); fp->fi_id = current_fileid++; fp->fi_had_conflict = false; + memset(fp->fi_fds, 0, sizeof(fp->fi_fds)); + memset(fp->fi_access, 0, sizeof(fp->fi_access)); spin_lock(&recall_lock); list_add(&fp->fi_hash, &file_hashtbl[hashval]); spin_unlock(&recall_lock); @@ -1971,57 +2212,6 @@ static inline int deny_valid(u32 x) } /* - * We store the NONE, READ, WRITE, and BOTH bits separately in the - * st_{access,deny}_bmap field of the stateid, in order to track not - * only what share bits are currently in force, but also what - * combinations of share bits previous opens have used. This allows us - * to enforce the recommendation of rfc 3530 14.2.19 that the server - * return an error if the client attempt to downgrade to a combination - * of share bits not explicable by closing some of its previous opens. - * - * XXX: This enforcement is actually incomplete, since we don't keep - * track of access/deny bit combinations; so, e.g., we allow: - * - * OPEN allow read, deny write - * OPEN allow both, deny none - * DOWNGRADE allow read, deny none - * - * which we should reject. - */ -static void -set_access(unsigned int *access, unsigned long bmap) { - int i; - - *access = 0; - for (i = 1; i < 4; i++) { - if (test_bit(i, &bmap)) - *access |= i; - } -} - -static void -set_deny(unsigned int *deny, unsigned long bmap) { - int i; - - *deny = 0; - for (i = 0; i < 4; i++) { - if (test_bit(i, &bmap)) - *deny |= i ; - } -} - -static int -test_share(struct nfs4_stateid *stp, struct nfsd4_open *open) { - unsigned int access, deny; - - set_access(&access, stp->st_access_bmap); - set_deny(&deny, stp->st_deny_bmap); - if ((access & open->op_share_deny) || (deny & open->op_share_access)) - return 0; - return 1; -} - -/* * Called to check deny when READ with all zero stateid or * WRITE with all zero or all one stateid */ @@ -2052,14 +2242,12 @@ out: } static inline void -nfs4_file_downgrade(struct file *filp, unsigned int share_access) +nfs4_file_downgrade(struct nfs4_file *fp, unsigned int share_access) { - if (share_access & NFS4_SHARE_ACCESS_WRITE) { - drop_file_write_access(filp); - spin_lock(&filp->f_lock); - filp->f_mode = (filp->f_mode | FMODE_READ) & ~FMODE_WRITE; - spin_unlock(&filp->f_lock); - } + if (share_access & NFS4_SHARE_ACCESS_WRITE) + nfs4_file_put_access(fp, O_WRONLY); + if (share_access & NFS4_SHARE_ACCESS_READ) + nfs4_file_put_access(fp, O_RDONLY); } /* @@ -2122,22 +2310,6 @@ void nfsd_release_deleg_cb(struct file_lock *fl) } /* - * Set the delegation file_lock back pointer. - * - * Called from setlease() with lock_kernel() held. - */ -static -void nfsd_copy_lock_deleg_cb(struct file_lock *new, struct file_lock *fl) -{ - struct nfs4_delegation *dp = (struct nfs4_delegation *)new->fl_owner; - - dprintk("NFSD: nfsd_copy_lock_deleg_cb: new fl %p dp %p\n", new, dp); - if (!dp) - return; - dp->dl_flock = new; -} - -/* * Called from setlease() with lock_kernel() held */ static @@ -2167,7 +2339,6 @@ int nfsd_change_deleg_cb(struct file_lock **onlist, int arg) static const struct lock_manager_operations nfsd_lease_mng_ops = { .fl_break = nfsd_break_deleg_cb, .fl_release_private = nfsd_release_deleg_cb, - .fl_copy_lock = nfsd_copy_lock_deleg_cb, .fl_mylease = nfsd_same_client_deleg_cb, .fl_change = nfsd_change_deleg_cb, }; @@ -2255,6 +2426,13 @@ find_delegation_file(struct nfs4_file *fp, stateid_t *stid) return NULL; } +int share_access_to_flags(u32 share_access) +{ + share_access &= ~NFS4_SHARE_WANT_MASK; + + return share_access == NFS4_SHARE_ACCESS_READ ? RD_STATE : WR_STATE; +} + static __be32 nfs4_check_deleg(struct nfs4_file *fp, struct nfsd4_open *open, struct nfs4_delegation **dp) @@ -2265,8 +2443,7 @@ nfs4_check_deleg(struct nfs4_file *fp, struct nfsd4_open *open, *dp = find_delegation_file(fp, &open->op_delegate_stateid); if (*dp == NULL) goto out; - flags = open->op_share_access == NFS4_SHARE_ACCESS_READ ? - RD_STATE : WR_STATE; + flags = share_access_to_flags(open->op_share_access); status = nfs4_check_delegmode(*dp, flags); if (status) *dp = NULL; @@ -2308,30 +2485,53 @@ nfs4_alloc_stateid(void) return kmem_cache_alloc(stateid_slab, GFP_KERNEL); } +static inline int nfs4_access_to_access(u32 nfs4_access) +{ + int flags = 0; + + if (nfs4_access & NFS4_SHARE_ACCESS_READ) + flags |= NFSD_MAY_READ; + if (nfs4_access & NFS4_SHARE_ACCESS_WRITE) + flags |= NFSD_MAY_WRITE; + return flags; +} + +static __be32 nfs4_get_vfs_file(struct svc_rqst *rqstp, struct nfs4_file +*fp, struct svc_fh *cur_fh, u32 nfs4_access) +{ + __be32 status; + int oflag = nfs4_access_to_omode(nfs4_access); + int access = nfs4_access_to_access(nfs4_access); + + if (!fp->fi_fds[oflag]) { + status = nfsd_open(rqstp, cur_fh, S_IFREG, access, + &fp->fi_fds[oflag]); + if (status == nfserr_dropit) + status = nfserr_jukebox; + if (status) + return status; + } + nfs4_file_get_access(fp, oflag); + + return nfs_ok; +} + static __be32 nfs4_new_open(struct svc_rqst *rqstp, struct nfs4_stateid **stpp, - struct nfs4_delegation *dp, - struct svc_fh *cur_fh, int flags) + struct nfs4_file *fp, struct svc_fh *cur_fh, + struct nfsd4_open *open) { struct nfs4_stateid *stp; + __be32 status; stp = nfs4_alloc_stateid(); if (stp == NULL) return nfserr_resource; - if (dp) { - get_file(dp->dl_vfs_file); - stp->st_vfs_file = dp->dl_vfs_file; - } else { - __be32 status; - status = nfsd_open(rqstp, cur_fh, S_IFREG, flags, - &stp->st_vfs_file); - if (status) { - if (status == nfserr_dropit) - status = nfserr_jukebox; - kmem_cache_free(stateid_slab, stp); - return status; - } + status = nfs4_get_vfs_file(rqstp, fp, cur_fh, open->op_share_access); + if (status) { + kmem_cache_free(stateid_slab, stp); + return status; } *stpp = stp; return 0; @@ -2353,35 +2553,28 @@ nfsd4_truncate(struct svc_rqst *rqstp, struct svc_fh *fh, } static __be32 -nfs4_upgrade_open(struct svc_rqst *rqstp, struct svc_fh *cur_fh, struct nfs4_stateid *stp, struct nfsd4_open *open) +nfs4_upgrade_open(struct svc_rqst *rqstp, struct nfs4_file *fp, struct svc_fh *cur_fh, struct nfs4_stateid *stp, struct nfsd4_open *open) { - struct file *filp = stp->st_vfs_file; - struct inode *inode = filp->f_path.dentry->d_inode; - unsigned int share_access, new_writer; + u32 op_share_access = open->op_share_access & ~NFS4_SHARE_WANT_MASK; + bool new_access; __be32 status; - set_access(&share_access, stp->st_access_bmap); - new_writer = (~share_access) & open->op_share_access - & NFS4_SHARE_ACCESS_WRITE; - - if (new_writer) { - int err = get_write_access(inode); - if (err) - return nfserrno(err); - err = mnt_want_write(cur_fh->fh_export->ex_path.mnt); - if (err) - return nfserrno(err); - file_take_write(filp); + new_access = !test_bit(op_share_access, &stp->st_access_bmap); + if (new_access) { + status = nfs4_get_vfs_file(rqstp, fp, cur_fh, op_share_access); + if (status) + return status; } status = nfsd4_truncate(rqstp, cur_fh, open); if (status) { - if (new_writer) - put_write_access(inode); + if (new_access) { + int oflag = nfs4_access_to_omode(new_access); + nfs4_file_put_access(fp, oflag); + } return status; } /* remember the open */ - filp->f_mode |= open->op_share_access; - __set_bit(open->op_share_access, &stp->st_access_bmap); + __set_bit(op_share_access, &stp->st_access_bmap); __set_bit(open->op_share_deny, &stp->st_deny_bmap); return nfs_ok; @@ -2404,7 +2597,7 @@ nfs4_open_delegation(struct svc_fh *fh, struct nfsd4_open *open, struct nfs4_sta struct nfs4_delegation *dp; struct nfs4_stateowner *sop = stp->st_stateowner; int cb_up = atomic_read(&sop->so_client->cl_cb_set); - struct file_lock fl, *flp = &fl; + struct file_lock *fl; int status, flag = 0; flag = NFS4_OPEN_DELEGATE_NONE; @@ -2438,20 +2631,28 @@ nfs4_open_delegation(struct svc_fh *fh, struct nfsd4_open *open, struct nfs4_sta flag = NFS4_OPEN_DELEGATE_NONE; goto out; } - locks_init_lock(&fl); - fl.fl_lmops = &nfsd_lease_mng_ops; - fl.fl_flags = FL_LEASE; - fl.fl_type = flag == NFS4_OPEN_DELEGATE_READ? F_RDLCK: F_WRLCK; - fl.fl_end = OFFSET_MAX; - fl.fl_owner = (fl_owner_t)dp; - fl.fl_file = stp->st_vfs_file; - fl.fl_pid = current->tgid; + status = -ENOMEM; + fl = locks_alloc_lock(); + if (!fl) + goto out; + locks_init_lock(fl); + fl->fl_lmops = &nfsd_lease_mng_ops; + fl->fl_flags = FL_LEASE; + fl->fl_type = flag == NFS4_OPEN_DELEGATE_READ? F_RDLCK: F_WRLCK; + fl->fl_end = OFFSET_MAX; + fl->fl_owner = (fl_owner_t)dp; + fl->fl_file = find_readable_file(stp->st_file); + BUG_ON(!fl->fl_file); + fl->fl_pid = current->tgid; + dp->dl_flock = fl; /* vfs_setlease checks to see if delegation should be handed out. * the lock_manager callbacks fl_mylease and fl_change are used */ - if ((status = vfs_setlease(stp->st_vfs_file, fl.fl_type, &flp))) { + if ((status = vfs_setlease(fl->fl_file, fl->fl_type, &fl))) { dprintk("NFSD: setlease failed [%d], no delegation\n", status); + dp->dl_flock = NULL; + locks_free_lock(fl); unhash_delegation(dp); flag = NFS4_OPEN_DELEGATE_NONE; goto out; @@ -2514,18 +2715,12 @@ nfsd4_process_open2(struct svc_rqst *rqstp, struct svc_fh *current_fh, struct nf */ if (stp) { /* Stateid was found, this is an OPEN upgrade */ - status = nfs4_upgrade_open(rqstp, current_fh, stp, open); + status = nfs4_upgrade_open(rqstp, fp, current_fh, stp, open); if (status) goto out; update_stateid(&stp->st_stateid); } else { - /* Stateid was not found, this is a new OPEN */ - int flags = 0; - if (open->op_share_access & NFS4_SHARE_ACCESS_READ) - flags |= NFSD_MAY_READ; - if (open->op_share_access & NFS4_SHARE_ACCESS_WRITE) - flags |= NFSD_MAY_WRITE; - status = nfs4_new_open(rqstp, &stp, dp, current_fh, flags); + status = nfs4_new_open(rqstp, &stp, fp, current_fh, open); if (status) goto out; init_stateid(stp, fp, open); @@ -2727,7 +2922,7 @@ search_close_lru(u32 st_id, int flags) static inline int nfs4_check_fh(struct svc_fh *fhp, struct nfs4_stateid *stp) { - return fhp->fh_dentry->d_inode != stp->st_vfs_file->f_path.dentry->d_inode; + return fhp->fh_dentry->d_inode != stp->st_file->fi_inode; } static int @@ -2760,6 +2955,9 @@ __be32 nfs4_check_openmode(struct nfs4_stateid *stp, int flags) { __be32 status = nfserr_openmode; + /* For lock stateid's, we test the parent open, not the lock: */ + if (stp->st_openstp) + stp = stp->st_openstp; if ((flags & WR_STATE) && (!access_permit_write(stp->st_access_bmap))) goto out; if ((flags & RD_STATE) && (!access_permit_read(stp->st_access_bmap))) @@ -2858,7 +3056,11 @@ nfs4_preprocess_stateid_op(struct nfsd4_compound_state *cstate, if (STALE_STATEID(stateid)) goto out; - status = nfserr_bad_stateid; + /* + * We assume that any stateid that has the current boot time, + * but that we can't find, is expired: + */ + status = nfserr_expired; if (is_delegation_stateid(stateid)) { dp = find_delegation_stateid(ino, stateid); if (!dp) @@ -2872,11 +3074,13 @@ nfs4_preprocess_stateid_op(struct nfsd4_compound_state *cstate, goto out; renew_client(dp->dl_client); if (filpp) - *filpp = dp->dl_vfs_file; + *filpp = find_readable_file(dp->dl_file); + BUG_ON(!*filpp); } else { /* open or lock stateid */ stp = find_stateid(stateid, flags); if (!stp) goto out; + status = nfserr_bad_stateid; if (nfs4_check_fh(current_fh, stp)) goto out; if (!stp->st_stateowner->so_confirmed) @@ -2889,8 +3093,12 @@ nfs4_preprocess_stateid_op(struct nfsd4_compound_state *cstate, if (status) goto out; renew_client(stp->st_stateowner->so_client); - if (filpp) - *filpp = stp->st_vfs_file; + if (filpp) { + if (flags & RD_STATE) + *filpp = find_readable_file(stp->st_file); + else + *filpp = find_writeable_file(stp->st_file); + } } status = nfs_ok; out: @@ -2947,8 +3155,9 @@ nfs4_preprocess_seqid_op(struct nfsd4_compound_state *cstate, u32 seqid, * a replayed close: */ sop = search_close_lru(stateid->si_stateownerid, flags); + /* It's not stale; let's assume it's expired: */ if (sop == NULL) - return nfserr_bad_stateid; + return nfserr_expired; *sopp = sop; goto check_replay; } @@ -3126,8 +3335,7 @@ nfsd4_open_downgrade(struct svc_rqst *rqstp, goto out; } set_access(&share_access, stp->st_access_bmap); - nfs4_file_downgrade(stp->st_vfs_file, - share_access & ~od->od_share_access); + nfs4_file_downgrade(stp->st_file, share_access & ~od->od_share_access); reset_union_bmap_access(od->od_share_access, &stp->st_access_bmap); reset_union_bmap_deny(od->od_share_deny, &stp->st_deny_bmap); @@ -3214,6 +3422,7 @@ nfsd4_delegreturn(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, status = nfserr_bad_stateid; if (!is_delegation_stateid(stateid)) goto out; + status = nfserr_expired; dp = find_delegation_stateid(inode, stateid); if (!dp) goto out; @@ -3346,11 +3555,9 @@ static inline void nfs4_set_lock_denied(struct file_lock *fl, struct nfsd4_lock_denied *deny) { struct nfs4_stateowner *sop; - unsigned int hval; if (fl->fl_lmops == &nfsd_posix_mng_ops) { sop = (struct nfs4_stateowner *) fl->fl_owner; - hval = lockownerid_hashval(sop->so_id); kref_get(&sop->so_ref); deny->ld_sop = sop; deny->ld_clientid = sop->so_client->cl_clientid; @@ -3446,8 +3653,6 @@ alloc_init_lock_stateid(struct nfs4_stateowner *sop, struct nfs4_file *fp, struc stp->st_stateid.si_stateownerid = sop->so_id; stp->st_stateid.si_fileid = fp->fi_id; stp->st_stateid.si_generation = 0; - stp->st_vfs_file = open_stp->st_vfs_file; /* FIXME refcount?? */ - stp->st_access_bmap = open_stp->st_access_bmap; stp->st_deny_bmap = open_stp->st_deny_bmap; stp->st_openstp = open_stp; @@ -3472,7 +3677,8 @@ nfsd4_lock(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, struct nfs4_stateowner *open_sop = NULL; struct nfs4_stateowner *lock_sop = NULL; struct nfs4_stateid *lock_stp; - struct file *filp; + struct nfs4_file *fp; + struct file *filp = NULL; struct file_lock file_lock; struct file_lock conflock; __be32 status = 0; @@ -3502,7 +3708,6 @@ nfsd4_lock(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, * lock stateid. */ struct nfs4_stateid *open_stp = NULL; - struct nfs4_file *fp; status = nfserr_stale_clientid; if (!nfsd4_has_session(cstate) && @@ -3545,9 +3750,9 @@ nfsd4_lock(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, if (status) goto out; lock_sop = lock->lk_replay_owner; + fp = lock_stp->st_file; } /* lock->lk_replay_owner and lock_stp have been created or found */ - filp = lock_stp->st_vfs_file; status = nfserr_grace; if (locks_in_grace() && !lock->lk_reclaim) @@ -3560,11 +3765,19 @@ nfsd4_lock(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, switch (lock->lk_type) { case NFS4_READ_LT: case NFS4_READW_LT: + if (find_readable_file(lock_stp->st_file)) { + nfs4_get_vfs_file(rqstp, fp, &cstate->current_fh, NFS4_SHARE_ACCESS_READ); + filp = find_readable_file(lock_stp->st_file); + } file_lock.fl_type = F_RDLCK; cmd = F_SETLK; break; case NFS4_WRITE_LT: case NFS4_WRITEW_LT: + if (find_writeable_file(lock_stp->st_file)) { + nfs4_get_vfs_file(rqstp, fp, &cstate->current_fh, NFS4_SHARE_ACCESS_WRITE); + filp = find_writeable_file(lock_stp->st_file); + } file_lock.fl_type = F_WRLCK; cmd = F_SETLK; break; @@ -3572,6 +3785,10 @@ nfsd4_lock(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, status = nfserr_inval; goto out; } + if (!filp) { + status = nfserr_openmode; + goto out; + } file_lock.fl_owner = (fl_owner_t)lock_sop; file_lock.fl_pid = current->tgid; file_lock.fl_file = filp; @@ -3740,7 +3957,11 @@ nfsd4_locku(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, &locku->lu_stateowner, &stp, NULL))) goto out; - filp = stp->st_vfs_file; + filp = find_any_file(stp->st_file); + if (!filp) { + status = nfserr_lock_range; + goto out; + } BUG_ON(!filp); locks_init_lock(&file_lock); file_lock.fl_type = F_UNLCK; @@ -3787,13 +4008,13 @@ out_nfserr: * 0: no locks held by lockowner */ static int -check_for_locks(struct file *filp, struct nfs4_stateowner *lowner) +check_for_locks(struct nfs4_file *filp, struct nfs4_stateowner *lowner) { struct file_lock **flpp; - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = filp->fi_inode; int status = 0; - lock_kernel(); + lock_flocks(); for (flpp = &inode->i_flock; *flpp != NULL; flpp = &(*flpp)->fl_next) { if ((*flpp)->fl_owner == (fl_owner_t)lowner) { status = 1; @@ -3801,7 +4022,7 @@ check_for_locks(struct file *filp, struct nfs4_stateowner *lowner) } } out: - unlock_kernel(); + unlock_flocks(); return status; } @@ -3841,7 +4062,7 @@ nfsd4_release_lockowner(struct svc_rqst *rqstp, continue; list_for_each_entry(stp, &sop->so_stateids, st_perstateowner) { - if (check_for_locks(stp->st_vfs_file, sop)) + if (check_for_locks(stp->st_file, sop)) goto out; /* Note: so_perclient unused for lockowners, * so it's OK to fool with here. */ @@ -4066,16 +4287,8 @@ out_free_laundry: int nfs4_state_start(void) { - int ret; - - if (nfs4_init) - return 0; nfsd4_load_reboot_recovery_data(); - ret = __nfs4_state_start(); - if (ret) - return ret; - nfs4_init = 1; - return 0; + return __nfs4_state_start(); } static void @@ -4110,7 +4323,6 @@ __nfs4_state_shutdown(void) } nfsd4_shutdown_recdir(); - nfs4_init = 0; } void |