diff options
Diffstat (limited to 'fs')
41 files changed, 3830 insertions, 2088 deletions
diff --git a/fs/coda/dir.c b/fs/coda/dir.c index 9ddf5ed62162..898a86dde8f5 100644 --- a/fs/coda/dir.c +++ b/fs/coda/dir.c @@ -470,7 +470,7 @@ int coda_readdir(struct file *coda_file, void *dirent, filldir_t filldir) ret = -ENOENT; if (!IS_DEADDIR(host_inode)) { - ret = host_file->f_op->readdir(host_file, filldir, dirent); + ret = host_file->f_op->readdir(host_file, dirent, filldir); file_accessed(host_file); } } diff --git a/fs/debugfs/inode.c b/fs/debugfs/inode.c index ec8896b264de..1d533a2ec3a6 100644 --- a/fs/debugfs/inode.c +++ b/fs/debugfs/inode.c @@ -368,6 +368,69 @@ void debugfs_remove(struct dentry *dentry) } EXPORT_SYMBOL_GPL(debugfs_remove); +/** + * debugfs_rename - rename a file/directory in the debugfs filesystem + * @old_dir: a pointer to the parent dentry for the renamed object. This + * should be a directory dentry. + * @old_dentry: dentry of an object to be renamed. + * @new_dir: a pointer to the parent dentry where the object should be + * moved. This should be a directory dentry. + * @new_name: a pointer to a string containing the target name. + * + * This function renames a file/directory in debugfs. The target must not + * exist for rename to succeed. + * + * This function will return a pointer to old_dentry (which is updated to + * reflect renaming) if it succeeds. If an error occurs, %NULL will be + * returned. + * + * If debugfs is not enabled in the kernel, the value -%ENODEV will be + * returned. + */ +struct dentry *debugfs_rename(struct dentry *old_dir, struct dentry *old_dentry, + struct dentry *new_dir, const char *new_name) +{ + int error; + struct dentry *dentry = NULL, *trap; + const char *old_name; + + trap = lock_rename(new_dir, old_dir); + /* Source or destination directories don't exist? */ + if (!old_dir->d_inode || !new_dir->d_inode) + goto exit; + /* Source does not exist, cyclic rename, or mountpoint? */ + if (!old_dentry->d_inode || old_dentry == trap || + d_mountpoint(old_dentry)) + goto exit; + dentry = lookup_one_len(new_name, new_dir, strlen(new_name)); + /* Lookup failed, cyclic rename or target exists? */ + if (IS_ERR(dentry) || dentry == trap || dentry->d_inode) + goto exit; + + old_name = fsnotify_oldname_init(old_dentry->d_name.name); + + error = simple_rename(old_dir->d_inode, old_dentry, new_dir->d_inode, + dentry); + if (error) { + fsnotify_oldname_free(old_name); + goto exit; + } + d_move(old_dentry, dentry); + fsnotify_move(old_dir->d_inode, new_dir->d_inode, old_name, + old_dentry->d_name.name, S_ISDIR(old_dentry->d_inode->i_mode), + NULL, old_dentry->d_inode); + fsnotify_oldname_free(old_name); + unlock_rename(new_dir, old_dir); + dput(dentry); + return old_dentry; +exit: + if (dentry && !IS_ERR(dentry)) + dput(dentry); + unlock_rename(new_dir, old_dir); + return NULL; +} +EXPORT_SYMBOL_GPL(debugfs_rename); + static decl_subsys(debug, NULL, NULL); static int __init debugfs_init(void) diff --git a/fs/ecryptfs/main.c b/fs/ecryptfs/main.c index 606128f5c927..02ca6f1e55d7 100644 --- a/fs/ecryptfs/main.c +++ b/fs/ecryptfs/main.c @@ -840,8 +840,6 @@ static int __init ecryptfs_init(void) goto out; } kobj_set_kset_s(&ecryptfs_subsys, fs_subsys); - sysfs_attr_version.attr.owner = THIS_MODULE; - sysfs_attr_version_str.attr.owner = THIS_MODULE; rc = do_sysfs_registration(); if (rc) { printk(KERN_ERR "sysfs registration failed\n"); diff --git a/fs/lockd/host.c b/fs/lockd/host.c index 96070bff93fc..572601e98dcd 100644 --- a/fs/lockd/host.c +++ b/fs/lockd/host.c @@ -44,9 +44,8 @@ static struct nsm_handle * nsm_find(const struct sockaddr_in *sin, */ static struct nlm_host * nlm_lookup_host(int server, const struct sockaddr_in *sin, - int proto, int version, - const char *hostname, - int hostname_len) + int proto, int version, const char *hostname, + int hostname_len, const struct sockaddr_in *ssin) { struct hlist_head *chain; struct hlist_node *pos; @@ -54,7 +53,9 @@ nlm_lookup_host(int server, const struct sockaddr_in *sin, struct nsm_handle *nsm = NULL; int hash; - dprintk("lockd: nlm_lookup_host(%u.%u.%u.%u, p=%d, v=%d, my role=%s, name=%.*s)\n", + dprintk("lockd: nlm_lookup_host("NIPQUAD_FMT"->"NIPQUAD_FMT + ", p=%d, v=%d, my role=%s, name=%.*s)\n", + NIPQUAD(ssin->sin_addr.s_addr), NIPQUAD(sin->sin_addr.s_addr), proto, version, server? "server" : "client", hostname_len, @@ -91,6 +92,8 @@ nlm_lookup_host(int server, const struct sockaddr_in *sin, continue; if (host->h_server != server) continue; + if (!nlm_cmp_addr(&host->h_saddr, ssin)) + continue; /* Move to head of hash chain. */ hlist_del(&host->h_hash); @@ -118,6 +121,7 @@ nlm_lookup_host(int server, const struct sockaddr_in *sin, host->h_name = nsm->sm_name; host->h_addr = *sin; host->h_addr.sin_port = 0; /* ouch! */ + host->h_saddr = *ssin; host->h_version = version; host->h_proto = proto; host->h_rpcclnt = NULL; @@ -161,15 +165,9 @@ nlm_destroy_host(struct nlm_host *host) */ nsm_unmonitor(host); - if ((clnt = host->h_rpcclnt) != NULL) { - if (atomic_read(&clnt->cl_users)) { - printk(KERN_WARNING - "lockd: active RPC handle\n"); - clnt->cl_dead = 1; - } else { - rpc_destroy_client(host->h_rpcclnt); - } - } + clnt = host->h_rpcclnt; + if (clnt != NULL) + rpc_shutdown_client(clnt); kfree(host); } @@ -180,8 +178,10 @@ struct nlm_host * nlmclnt_lookup_host(const struct sockaddr_in *sin, int proto, int version, const char *hostname, int hostname_len) { + struct sockaddr_in ssin = {0}; + return nlm_lookup_host(0, sin, proto, version, - hostname, hostname_len); + hostname, hostname_len, &ssin); } /* @@ -191,9 +191,12 @@ struct nlm_host * nlmsvc_lookup_host(struct svc_rqst *rqstp, const char *hostname, int hostname_len) { + struct sockaddr_in ssin = {0}; + + ssin.sin_addr = rqstp->rq_daddr.addr; return nlm_lookup_host(1, svc_addr_in(rqstp), rqstp->rq_prot, rqstp->rq_vers, - hostname, hostname_len); + hostname, hostname_len, &ssin); } /* @@ -204,8 +207,9 @@ nlm_bind_host(struct nlm_host *host) { struct rpc_clnt *clnt; - dprintk("lockd: nlm_bind_host(%08x)\n", - (unsigned)ntohl(host->h_addr.sin_addr.s_addr)); + dprintk("lockd: nlm_bind_host("NIPQUAD_FMT"->"NIPQUAD_FMT")\n", + NIPQUAD(host->h_saddr.sin_addr), + NIPQUAD(host->h_addr.sin_addr)); /* Lock host handle */ mutex_lock(&host->h_mutex); @@ -232,6 +236,7 @@ nlm_bind_host(struct nlm_host *host) .protocol = host->h_proto, .address = (struct sockaddr *)&host->h_addr, .addrsize = sizeof(host->h_addr), + .saddress = (struct sockaddr *)&host->h_saddr, .timeout = &timeparms, .servername = host->h_name, .program = &nlm_program, diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index 2102e2d0134d..3353ed8421a7 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -61,6 +61,7 @@ nsm_mon_unmon(struct nsm_handle *nsm, u32 proc, struct nsm_res *res) status); else status = 0; + rpc_shutdown_client(clnt); out: return status; } @@ -138,7 +139,6 @@ nsm_create(void) .program = &nsm_program, .version = SM_VERSION, .authflavor = RPC_AUTH_NULL, - .flags = (RPC_CLNT_CREATE_ONESHOT), }; return rpc_create(&args); diff --git a/fs/lockd/svc.c b/fs/lockd/svc.c index 126b1bf02c0e..26809325469c 100644 --- a/fs/lockd/svc.c +++ b/fs/lockd/svc.c @@ -123,9 +123,6 @@ lockd(struct svc_rqst *rqstp) /* Process request with signals blocked, but allow SIGKILL. */ allow_signal(SIGKILL); - /* kick rpciod */ - rpciod_up(); - dprintk("NFS locking service started (ver " LOCKD_VERSION ").\n"); if (!nlm_timeout) @@ -202,9 +199,6 @@ lockd(struct svc_rqst *rqstp) /* Exit the RPC thread */ svc_exit_thread(rqstp); - /* release rpciod */ - rpciod_down(); - /* Release module */ unlock_kernel(); module_put_and_exit(0); diff --git a/fs/nfs/Makefile b/fs/nfs/Makefile index f4580b44eef4..b55cb236cf74 100644 --- a/fs/nfs/Makefile +++ b/fs/nfs/Makefile @@ -6,8 +6,8 @@ obj-$(CONFIG_NFS_FS) += nfs.o nfs-y := client.o dir.o file.o getroot.o inode.o super.o nfs2xdr.o \ pagelist.o proc.o read.o symlink.o unlink.o \ - write.o namespace.o -nfs-$(CONFIG_ROOT_NFS) += nfsroot.o mount_clnt.o + write.o namespace.o mount_clnt.o +nfs-$(CONFIG_ROOT_NFS) += nfsroot.o nfs-$(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 \ diff --git a/fs/nfs/client.c b/fs/nfs/client.c index 881fa4900923..ccb455053ee4 100644 --- a/fs/nfs/client.c +++ b/fs/nfs/client.c @@ -102,19 +102,10 @@ static struct nfs_client *nfs_alloc_client(const char *hostname, int nfsversion) { struct nfs_client *clp; - int error; if ((clp = kzalloc(sizeof(*clp), GFP_KERNEL)) == NULL) goto error_0; - error = rpciod_up(); - if (error < 0) { - dprintk("%s: couldn't start rpciod! Error = %d\n", - __FUNCTION__, error); - goto error_1; - } - __set_bit(NFS_CS_RPCIOD, &clp->cl_res_state); - if (nfsversion == 4) { if (nfs_callback_up() < 0) goto error_2; @@ -139,8 +130,6 @@ static struct nfs_client *nfs_alloc_client(const char *hostname, #ifdef CONFIG_NFS_V4 init_rwsem(&clp->cl_sem); INIT_LIST_HEAD(&clp->cl_delegations); - INIT_LIST_HEAD(&clp->cl_state_owners); - INIT_LIST_HEAD(&clp->cl_unused); spin_lock_init(&clp->cl_lock); INIT_DELAYED_WORK(&clp->cl_renewd, nfs4_renew_state); rpc_init_wait_queue(&clp->cl_rpcwaitq, "NFS client"); @@ -154,9 +143,6 @@ error_3: if (__test_and_clear_bit(NFS_CS_CALLBACK, &clp->cl_res_state)) nfs_callback_down(); error_2: - rpciod_down(); - __clear_bit(NFS_CS_RPCIOD, &clp->cl_res_state); -error_1: kfree(clp); error_0: return NULL; @@ -167,16 +153,7 @@ static void nfs4_shutdown_client(struct nfs_client *clp) #ifdef CONFIG_NFS_V4 if (__test_and_clear_bit(NFS_CS_RENEWD, &clp->cl_res_state)) nfs4_kill_renewd(clp); - while (!list_empty(&clp->cl_unused)) { - struct nfs4_state_owner *sp; - - sp = list_entry(clp->cl_unused.next, - struct nfs4_state_owner, - so_list); - list_del(&sp->so_list); - kfree(sp); - } - BUG_ON(!list_empty(&clp->cl_state_owners)); + BUG_ON(!RB_EMPTY_ROOT(&clp->cl_state_owners)); if (__test_and_clear_bit(NFS_CS_IDMAP, &clp->cl_res_state)) nfs_idmap_delete(clp); #endif @@ -198,9 +175,6 @@ static void nfs_free_client(struct nfs_client *clp) if (__test_and_clear_bit(NFS_CS_CALLBACK, &clp->cl_res_state)) nfs_callback_down(); - if (__test_and_clear_bit(NFS_CS_RPCIOD, &clp->cl_res_state)) - rpciod_down(); - kfree(clp->cl_hostname); kfree(clp); diff --git a/fs/nfs/delegation.c b/fs/nfs/delegation.c index 7f37d1bea83f..20ac403469a0 100644 --- a/fs/nfs/delegation.c +++ b/fs/nfs/delegation.c @@ -27,6 +27,13 @@ static void nfs_free_delegation(struct nfs_delegation *delegation) kfree(delegation); } +static void nfs_free_delegation_callback(struct rcu_head *head) +{ + struct nfs_delegation *delegation = container_of(head, struct nfs_delegation, rcu); + + nfs_free_delegation(delegation); +} + static int nfs_delegation_claim_locks(struct nfs_open_context *ctx, struct nfs4_state *state) { struct inode *inode = state->inode; @@ -57,7 +64,7 @@ out_err: return status; } -static void nfs_delegation_claim_opens(struct inode *inode) +static void nfs_delegation_claim_opens(struct inode *inode, const nfs4_stateid *stateid) { struct nfs_inode *nfsi = NFS_I(inode); struct nfs_open_context *ctx; @@ -72,9 +79,11 @@ again: continue; if (!test_bit(NFS_DELEGATED_STATE, &state->flags)) continue; + if (memcmp(state->stateid.data, stateid->data, sizeof(state->stateid.data)) != 0) + continue; get_nfs_open_context(ctx); spin_unlock(&inode->i_lock); - err = nfs4_open_delegation_recall(ctx->dentry, state); + err = nfs4_open_delegation_recall(ctx, state, stateid); if (err >= 0) err = nfs_delegation_claim_locks(ctx, state); put_nfs_open_context(ctx); @@ -115,10 +124,6 @@ int nfs_inode_set_delegation(struct inode *inode, struct rpc_cred *cred, struct struct nfs_delegation *delegation; int status = 0; - /* Ensure we first revalidate the attributes and page cache! */ - if ((nfsi->cache_validity & (NFS_INO_REVAL_PAGECACHE|NFS_INO_INVALID_ATTR))) - __nfs_revalidate_inode(NFS_SERVER(inode), inode); - delegation = kmalloc(sizeof(*delegation), GFP_KERNEL); if (delegation == NULL) return -ENOMEM; @@ -131,10 +136,10 @@ int nfs_inode_set_delegation(struct inode *inode, struct rpc_cred *cred, struct delegation->inode = inode; spin_lock(&clp->cl_lock); - if (nfsi->delegation == NULL) { - list_add(&delegation->super_list, &clp->cl_delegations); - nfsi->delegation = delegation; + if (rcu_dereference(nfsi->delegation) == NULL) { + list_add_rcu(&delegation->super_list, &clp->cl_delegations); nfsi->delegation_state = delegation->type; + rcu_assign_pointer(nfsi->delegation, delegation); delegation = NULL; } else { if (memcmp(&delegation->stateid, &nfsi->delegation->stateid, @@ -145,6 +150,12 @@ int nfs_inode_set_delegation(struct inode *inode, struct rpc_cred *cred, struct status = -EIO; } } + + /* Ensure we revalidate the attributes and page cache! */ + spin_lock(&inode->i_lock); + nfsi->cache_validity |= NFS_INO_REVAL_FORCED; + spin_unlock(&inode->i_lock); + spin_unlock(&clp->cl_lock); kfree(delegation); return status; @@ -155,7 +166,7 @@ static int nfs_do_return_delegation(struct inode *inode, struct nfs_delegation * int res = 0; res = nfs4_proc_delegreturn(inode, delegation->cred, &delegation->stateid); - nfs_free_delegation(delegation); + call_rcu(&delegation->rcu, nfs_free_delegation_callback); return res; } @@ -170,33 +181,55 @@ static void nfs_msync_inode(struct inode *inode) /* * Basic procedure for returning a delegation to the server */ -int __nfs_inode_return_delegation(struct inode *inode) +static int __nfs_inode_return_delegation(struct inode *inode, struct nfs_delegation *delegation) { struct nfs_client *clp = NFS_SERVER(inode)->nfs_client; struct nfs_inode *nfsi = NFS_I(inode); - struct nfs_delegation *delegation; - int res = 0; nfs_msync_inode(inode); down_read(&clp->cl_sem); /* Guard against new delegated open calls */ down_write(&nfsi->rwsem); - spin_lock(&clp->cl_lock); - delegation = nfsi->delegation; - if (delegation != NULL) { - list_del_init(&delegation->super_list); - nfsi->delegation = NULL; - nfsi->delegation_state = 0; - } - spin_unlock(&clp->cl_lock); - nfs_delegation_claim_opens(inode); + nfs_delegation_claim_opens(inode, &delegation->stateid); up_write(&nfsi->rwsem); up_read(&clp->cl_sem); nfs_msync_inode(inode); - if (delegation != NULL) - res = nfs_do_return_delegation(inode, delegation); - return res; + return nfs_do_return_delegation(inode, delegation); +} + +static struct nfs_delegation *nfs_detach_delegation_locked(struct nfs_inode *nfsi, const nfs4_stateid *stateid) +{ + struct nfs_delegation *delegation = rcu_dereference(nfsi->delegation); + + if (delegation == NULL) + goto nomatch; + if (stateid != NULL && memcmp(delegation->stateid.data, stateid->data, + sizeof(delegation->stateid.data)) != 0) + goto nomatch; + list_del_rcu(&delegation->super_list); + nfsi->delegation_state = 0; + rcu_assign_pointer(nfsi->delegation, NULL); + return delegation; +nomatch: + return NULL; +} + +int nfs_inode_return_delegation(struct inode *inode) +{ + struct nfs_client *clp = NFS_SERVER(inode)->nfs_client; + struct nfs_inode *nfsi = NFS_I(inode); + struct nfs_delegation *delegation; + int err = 0; + + if (rcu_dereference(nfsi->delegation) != NULL) { + spin_lock(&clp->cl_lock); + delegation = nfs_detach_delegation_locked(nfsi, NULL); + spin_unlock(&clp->cl_lock); + if (delegation != NULL) + err = __nfs_inode_return_delegation(inode, delegation); + } + return err; } /* @@ -211,19 +244,23 @@ void nfs_return_all_delegations(struct super_block *sb) if (clp == NULL) return; restart: - spin_lock(&clp->cl_lock); - list_for_each_entry(delegation, &clp->cl_delegations, super_list) { + rcu_read_lock(); + list_for_each_entry_rcu(delegation, &clp->cl_delegations, super_list) { if (delegation->inode->i_sb != sb) continue; inode = igrab(delegation->inode); if (inode == NULL) continue; + spin_lock(&clp->cl_lock); + delegation = nfs_detach_delegation_locked(NFS_I(inode), NULL); spin_unlock(&clp->cl_lock); - nfs_inode_return_delegation(inode); + rcu_read_unlock(); + if (delegation != NULL) + __nfs_inode_return_delegation(inode, delegation); iput(inode); goto restart; } - spin_unlock(&clp->cl_lock); + rcu_read_unlock(); } static int nfs_do_expire_all_delegations(void *ptr) @@ -234,22 +271,26 @@ static int nfs_do_expire_all_delegations(void *ptr) allow_signal(SIGKILL); restart: - spin_lock(&clp->cl_lock); if (test_bit(NFS4CLNT_STATE_RECOVER, &clp->cl_state) != 0) goto out; if (test_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state) == 0) goto out; - list_for_each_entry(delegation, &clp->cl_delegations, super_list) { + rcu_read_lock(); + list_for_each_entry_rcu(delegation, &clp->cl_delegations, super_list) { inode = igrab(delegation->inode); if (inode == NULL) continue; + spin_lock(&clp->cl_lock); + delegation = nfs_detach_delegation_locked(NFS_I(inode), NULL); spin_unlock(&clp->cl_lock); - nfs_inode_return_delegation(inode); + rcu_read_unlock(); + if (delegation) + __nfs_inode_return_delegation(inode, delegation); iput(inode); goto restart; } + rcu_read_unlock(); out: - spin_unlock(&clp->cl_lock); nfs_put_client(clp); module_put_and_exit(0); } @@ -280,17 +321,21 @@ void nfs_handle_cb_pathdown(struct nfs_client *clp) if (clp == NULL) return; restart: - spin_lock(&clp->cl_lock); - list_for_each_entry(delegation, &clp->cl_delegations, super_list) { + rcu_read_lock(); + list_for_each_entry_rcu(delegation, &clp->cl_delegations, super_list) { inode = igrab(delegation->inode); if (inode == NULL) continue; + spin_lock(&clp->cl_lock); + delegation = nfs_detach_delegation_locked(NFS_I(inode), NULL); spin_unlock(&clp->cl_lock); - nfs_inode_return_delegation(inode); + rcu_read_unlock(); + if (delegation != NULL) + __nfs_inode_return_delegation(inode, delegation); iput(inode); goto restart; } - spin_unlock(&clp->cl_lock); + rcu_read_unlock(); } struct recall_threadargs { @@ -316,21 +361,14 @@ static int recall_thread(void *data) down_read(&clp->cl_sem); down_write(&nfsi->rwsem); spin_lock(&clp->cl_lock); - delegation = nfsi->delegation; - if (delegation != NULL && memcmp(delegation->stateid.data, - args->stateid->data, - sizeof(delegation->stateid.data)) == 0) { - list_del_init(&delegation->super_list); - nfsi->delegation = NULL; - nfsi->delegation_state = 0; + delegation = nfs_detach_delegation_locked(nfsi, args->stateid); + if (delegation != NULL) args->result = 0; - } else { - delegation = NULL; + else args->result = -ENOENT; - } spin_unlock(&clp->cl_lock); complete(&args->started); - nfs_delegation_claim_opens(inode); + nfs_delegation_claim_opens(inode, args->stateid); up_write(&nfsi->rwsem); up_read(&clp->cl_sem); nfs_msync_inode(inode); @@ -371,14 +409,14 @@ struct inode *nfs_delegation_find_inode(struct nfs_client *clp, const struct nfs { struct nfs_delegation *delegation; struct inode *res = NULL; - spin_lock(&clp->cl_lock); - list_for_each_entry(delegation, &clp->cl_delegations, super_list) { + rcu_read_lock(); + list_for_each_entry_rcu(delegation, &clp->cl_delegations, super_list) { if (nfs_compare_fh(fhandle, &NFS_I(delegation->inode)->fh) == 0) { res = igrab(delegation->inode); break; } } - spin_unlock(&clp->cl_lock); + rcu_read_unlock(); return res; } @@ -388,10 +426,10 @@ struct inode *nfs_delegation_find_inode(struct nfs_client *clp, const struct nfs void nfs_delegation_mark_reclaim(struct nfs_client *clp) { struct nfs_delegation *delegation; - spin_lock(&clp->cl_lock); - list_for_each_entry(delegation, &clp->cl_delegations, super_list) + rcu_read_lock(); + list_for_each_entry_rcu(delegation, &clp->cl_delegations, super_list) delegation->flags |= NFS_DELEGATION_NEED_RECLAIM; - spin_unlock(&clp->cl_lock); + rcu_read_unlock(); } /* @@ -399,39 +437,35 @@ void nfs_delegation_mark_reclaim(struct nfs_client *clp) */ void nfs_delegation_reap_unclaimed(struct nfs_client *clp) { - struct nfs_delegation *delegation, *n; - LIST_HEAD(head); - spin_lock(&clp->cl_lock); - list_for_each_entry_safe(delegation, n, &clp->cl_delegations, super_list) { + struct nfs_delegation *delegation; +restart: + rcu_read_lock(); + list_for_each_entry_rcu(delegation, &clp->cl_delegations, super_list) { if ((delegation->flags & NFS_DELEGATION_NEED_RECLAIM) == 0) continue; - list_move(&delegation->super_list, &head); - NFS_I(delegation->inode)->delegation = NULL; - NFS_I(delegation->inode)->delegation_state = 0; - } - spin_unlock(&clp->cl_lock); - while(!list_empty(&head)) { - delegation = list_entry(head.next, struct nfs_delegation, super_list); - list_del(&delegation->super_list); - nfs_free_delegation(delegation); + spin_lock(&clp->cl_lock); + delegation = nfs_detach_delegation_locked(NFS_I(delegation->inode), NULL); + spin_unlock(&clp->cl_lock); + rcu_read_unlock(); + if (delegation != NULL) + call_rcu(&delegation->rcu, nfs_free_delegation_callback); + goto restart; } + rcu_read_unlock(); } int nfs4_copy_delegation_stateid(nfs4_stateid *dst, struct inode *inode) { - struct nfs_client *clp = NFS_SERVER(inode)->nfs_client; struct nfs_inode *nfsi = NFS_I(inode); struct nfs_delegation *delegation; - int res = 0; + int ret = 0; - if (nfsi->delegation_state == 0) - return 0; - spin_lock(&clp->cl_lock); - delegation = nfsi->delegation; + rcu_read_lock(); + delegation = rcu_dereference(nfsi->delegation); if (delegation != NULL) { memcpy(dst->data, delegation->stateid.data, sizeof(dst->data)); - res = 1; + ret = 1; } - spin_unlock(&clp->cl_lock); - return res; + rcu_read_unlock(); + return ret; } diff --git a/fs/nfs/delegation.h b/fs/nfs/delegation.h index 2cfd4b24c7fe..5874ce7fdbae 100644 --- a/fs/nfs/delegation.h +++ b/fs/nfs/delegation.h @@ -22,11 +22,12 @@ struct nfs_delegation { long flags; loff_t maxsize; __u64 change_attr; + struct rcu_head rcu; }; int nfs_inode_set_delegation(struct inode *inode, struct rpc_cred *cred, struct nfs_openres *res); void nfs_inode_reclaim_delegation(struct inode *inode, struct rpc_cred *cred, struct nfs_openres *res); -int __nfs_inode_return_delegation(struct inode *inode); +int nfs_inode_return_delegation(struct inode *inode); int nfs_async_inode_return_delegation(struct inode *inode, const nfs4_stateid *stateid); struct inode *nfs_delegation_find_inode(struct nfs_client *clp, const struct nfs_fh *fhandle); @@ -39,27 +40,24 @@ void nfs_delegation_reap_unclaimed(struct nfs_client *clp); /* NFSv4 delegation-related procedures */ int nfs4_proc_delegreturn(struct inode *inode, struct rpc_cred *cred, const nfs4_stateid *stateid); -int nfs4_open_delegation_recall(struct dentry *dentry, struct nfs4_state *state); +int nfs4_open_delegation_recall(struct nfs_open_context *ctx, struct nfs4_state *state, const nfs4_stateid *stateid); int nfs4_lock_delegation_recall(struct nfs4_state *state, struct file_lock *fl); int nfs4_copy_delegation_stateid(nfs4_stateid *dst, struct inode *inode); static inline int nfs_have_delegation(struct inode *inode, int flags) { + struct nfs_delegation *delegation; + int ret = 0; + flags &= FMODE_READ|FMODE_WRITE; - smp_rmb(); - if ((NFS_I(inode)->delegation_state & flags) == flags) - return 1; - return 0; + rcu_read_lock(); + delegation = rcu_dereference(NFS_I(inode)->delegation); + if (delegation != NULL && (delegation->type & flags) == flags) + ret = 1; + rcu_read_unlock(); + return ret; } -static inline int nfs_inode_return_delegation(struct inode *inode) -{ - int err = 0; - - if (NFS_I(inode)->delegation != NULL) - err = __nfs_inode_return_delegation(inode); - return err; -} #else static inline int nfs_have_delegation(struct inode *inode, int flags) { diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c index c27258b5d3e1..322141f4ab48 100644 --- a/fs/nfs/dir.c +++ b/fs/nfs/dir.c @@ -897,14 +897,13 @@ int nfs_is_exclusive_create(struct inode *dir, struct nameidata *nd) return (nd->intent.open.flags & O_EXCL) != 0; } -static inline int nfs_reval_fsid(struct vfsmount *mnt, struct inode *dir, - struct nfs_fh *fh, struct nfs_fattr *fattr) +static inline int nfs_reval_fsid(struct inode *dir, const struct nfs_fattr *fattr) { struct nfs_server *server = NFS_SERVER(dir); if (!nfs_fsid_equal(&server->fsid, &fattr->fsid)) - /* Revalidate fsid on root dir */ - return __nfs_revalidate_inode(server, mnt->mnt_root->d_inode); + /* Revalidate fsid using the parent directory */ + return __nfs_revalidate_inode(server, dir); return 0; } @@ -946,7 +945,7 @@ static struct dentry *nfs_lookup(struct inode *dir, struct dentry * dentry, stru res = ERR_PTR(error); goto out_unlock; } - error = nfs_reval_fsid(nd->mnt, dir, &fhandle, &fattr); + error = nfs_reval_fsid(dir, &fattr); if (error < 0) { res = ERR_PTR(error); goto out_unlock; @@ -1244,7 +1243,7 @@ static int nfs_create(struct inode *dir, struct dentry *dentry, int mode, attr.ia_mode = mode; attr.ia_valid = ATTR_MODE; - if (nd && (nd->flags & LOOKUP_CREATE)) + if ((nd->flags & LOOKUP_CREATE) != 0) open_flags = nd->intent.open.flags; lock_kernel(); @@ -1535,7 +1534,7 @@ static int nfs_symlink(struct inode *dir, struct dentry *dentry, const char *sym lock_kernel(); - page = alloc_page(GFP_KERNEL); + page = alloc_page(GFP_HIGHUSER); if (!page) { unlock_kernel(); return -ENOMEM; @@ -1744,8 +1743,8 @@ int nfs_access_cache_shrinker(int nr_to_scan, gfp_t gfp_mask) struct nfs_inode *nfsi; struct nfs_access_entry *cache; - spin_lock(&nfs_access_lru_lock); restart: + spin_lock(&nfs_access_lru_lock); list_for_each_entry(nfsi, &nfs_access_lru_list, access_cache_inode_lru) { struct inode *inode; @@ -1770,6 +1769,7 @@ remove_lru_entry: clear_bit(NFS_INO_ACL_LRU_SET, &nfsi->flags); } spin_unlock(&inode->i_lock); + spin_unlock(&nfs_access_lru_lock); iput(inode); goto restart; } diff --git a/fs/nfs/direct.c b/fs/nfs/direct.c index 00eee87510fe..a5c82b6f3b45 100644 --- a/fs/nfs/direct.c +++ b/fs/nfs/direct.c @@ -266,7 +266,7 @@ static const struct rpc_call_ops nfs_read_direct_ops = { static ssize_t nfs_direct_read_schedule(struct nfs_direct_req *dreq, unsigned long user_addr, size_t count, loff_t pos) { struct nfs_open_context *ctx = dreq->ctx; - struct inode *inode = ctx->dentry->d_inode; + struct inode *inode = ctx->path.dentry->d_inode; size_t rsize = NFS_SERVER(inode)->rsize; unsigned int pgbase; int result; @@ -295,9 +295,14 @@ static ssize_t nfs_direct_read_schedule(struct nfs_direct_req *dreq, unsigned lo break; } if ((unsigned)result < data->npages) { - nfs_direct_release_pages(data->pagevec, result); - nfs_readdata_release(data); - break; + bytes = result * PAGE_SIZE; + if (bytes <= pgbase) { + nfs_direct_release_pages(data->pagevec, result); + nfs_readdata_release(data); + break; + } + bytes -= pgbase; + data->npages = result; } get_dreq(dreq); @@ -601,7 +606,7 @@ static const struct rpc_call_ops nfs_write_direct_ops = { static ssize_t nfs_direct_write_schedule(struct nfs_direct_req *dreq, unsigned long user_addr, size_t count, loff_t pos, int sync) { struct nfs_open_context *ctx = dreq->ctx; - struct inode *inode = ctx->dentry->d_inode; + struct inode *inode = ctx->path.dentry->d_inode; size_t wsize = NFS_SERVER(inode)->wsize; unsigned int pgbase; int result; @@ -630,9 +635,14 @@ static ssize_t nfs_direct_write_schedule(struct nfs_direct_req *dreq, unsigned l break; } if ((unsigned)result < data->npages) { - nfs_direct_release_pages(data->pagevec, result); - nfs_writedata_release(data); - break; + bytes = result * PAGE_SIZE; + if (bytes <= pgbase) { + nfs_direct_release_pages(data->pagevec, result); + nfs_writedata_release(data); + break; + } + bytes -= pgbase; + data->npages = result; } get_dreq(dreq); @@ -763,10 +773,8 @@ ssize_t nfs_file_direct_read(struct kiocb *iocb, const struct iovec *iov, (unsigned long) count, (long long) pos); if (nr_segs != 1) - return -EINVAL; - - if (count < 0) goto out; + retval = -EFAULT; if (!access_ok(VERIFY_WRITE, buf, count)) goto out; @@ -814,7 +822,7 @@ out: ssize_t nfs_file_direct_write(struct kiocb *iocb, const struct iovec *iov, unsigned long nr_segs, loff_t pos) { - ssize_t retval; + ssize_t retval = -EINVAL; struct file *file = iocb->ki_filp; struct address_space *mapping = file->f_mapping; /* XXX: temporary */ @@ -827,7 +835,7 @@ ssize_t nfs_file_direct_write(struct kiocb *iocb, const struct iovec *iov, (unsigned long) count, (long long) pos); if (nr_segs != 1) - return -EINVAL; + goto out; retval = generic_write_checks(file, &pos, &count, 0); if (retval) diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c index bd9f5a836592..3d9fccf4ef93 100644 --- a/fs/nfs/inode.c +++ b/fs/nfs/inode.c @@ -461,14 +461,14 @@ static struct nfs_open_context *alloc_nfs_open_context(struct vfsmount *mnt, str ctx = kmalloc(sizeof(*ctx), GFP_KERNEL); if (ctx != NULL) { - atomic_set(&ctx->count, 1); - ctx->dentry = dget(dentry); - ctx->vfsmnt = mntget(mnt); + ctx->path.dentry = dget(dentry); + ctx->path.mnt = mntget(mnt); ctx->cred = get_rpccred(cred); ctx->state = NULL; ctx->lockowner = current->files; ctx->error = 0; ctx->dir_cookie = 0; + kref_init(&ctx->kref); } return ctx; } @@ -476,27 +476,33 @@ static struct nfs_open_context *alloc_nfs_open_context(struct vfsmount *mnt, str struct nfs_open_context *get_nfs_open_context(struct nfs_open_context *ctx) { if (ctx != NULL) - atomic_inc(&ctx->count); + kref_get(&ctx->kref); return ctx; } -void put_nfs_open_context(struct nfs_open_context *ctx) +static void nfs_free_open_context(struct kref *kref) { - if (atomic_dec_and_test(&ctx->count)) { - if (!list_empty(&ctx->list)) { - struct inode *inode = ctx->dentry->d_inode; - spin_lock(&inode->i_lock); - list_del(&ctx->list); - spin_unlock(&inode->i_lock); - } - if (ctx->state != NULL) - nfs4_close_state(ctx->state, ctx->mode); - if (ctx->cred != NULL) - put_rpccred(ctx->cred); - dput(ctx->dentry); - mntput(ctx->vfsmnt); - kfree(ctx); + struct nfs_open_context *ctx = container_of(kref, + struct nfs_open_context, kref); + + if (!list_empty(&ctx->list)) { + struct inode *inode = ctx->path.dentry->d_inode; + spin_lock(&inode->i_lock); + list_del(&ctx->list); + spin_unlock(&inode->i_lock); } + if (ctx->state != NULL) + nfs4_close_state(&ctx->path, ctx->state, ctx->mode); + if (ctx->cred != NULL) + put_rpccred(ctx->cred); + dput(ctx->path.dentry); + mntput(ctx->path.mnt); + kfree(ctx); +} + +void put_nfs_open_context(struct nfs_open_context *ctx) +{ + kref_put(&ctx->kref, nfs_free_open_context); } /* @@ -961,8 +967,8 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr) goto out_changed; server = NFS_SERVER(inode); - /* Update the fsid if and only if this is the root directory */ - if (inode == inode->i_sb->s_root->d_inode + /* Update the fsid? */ + if (S_ISDIR(inode->i_mode) && !nfs_fsid_equal(&server->fsid, &fattr->fsid)) server->fsid = fattr->fsid; @@ -1066,8 +1072,10 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr) invalid &= ~NFS_INO_INVALID_DATA; if (data_stable) invalid &= ~(NFS_INO_INVALID_ATTR|NFS_INO_INVALID_ATIME|NFS_INO_REVAL_PAGECACHE); - if (!nfs_have_delegation(inode, FMODE_READ)) + if (!nfs_have_delegation(inode, FMODE_READ) || + (nfsi->cache_validity & NFS_INO_REVAL_FORCED)) nfsi->cache_validity |= invalid; + nfsi->cache_validity &= ~NFS_INO_REVAL_FORCED; return 0; out_changed: @@ -1103,27 +1111,10 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr) */ void nfs4_clear_inode(struct inode *inode) { - struct nfs_inode *nfsi = NFS_I(inode); - /* If we are holding a delegation, return it! */ nfs_inode_return_delegation(inode); /* First call standard NFS clear_inode() code */ nfs_clear_inode(inode); - /* Now clear out any remaining state */ - while (!list_empty(&nfsi->open_states)) { - struct nfs4_state *state; - - state = list_entry(nfsi->open_states.next, - struct nfs4_state, - inode_states); - dprintk("%s(%s/%Ld): found unclaimed NFSv4 state %p\n", - __FUNCTION__, - inode->i_sb->s_id, - (long long)NFS_FILEID(inode), - state); - BUG_ON(atomic_read(&state->count) != 1); - nfs4_close_state(state, state->state); - } } #endif @@ -1165,15 +1156,11 @@ static void init_once(void * foo, struct kmem_cache * cachep, unsigned long flag struct nfs_inode *nfsi = (struct nfs_inode *) foo; inode_init_once(&nfsi->vfs_inode); - spin_lock_init(&nfsi->req_lock); - INIT_LIST_HEAD(&nfsi->dirty); - INIT_LIST_HEAD(&nfsi->commit); INIT_LIST_HEAD(&nfsi->open_files); INIT_LIST_HEAD(&nfsi->access_cache_entry_lru); INIT_LIST_HEAD(&nfsi->access_cache_inode_lru); INIT_RADIX_TREE(&nfsi->nfs_page_tree, GFP_ATOMIC); atomic_set(&nfsi->data_updates, 0); - nfsi->ndirty = 0; nfsi->ncommit = 0; nfsi->npages = 0; nfs4_init_once(nfsi); diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h index ad2b40db1e65..76cf55d57101 100644 --- a/fs/nfs/internal.h +++ b/fs/nfs/internal.h @@ -183,9 +183,9 @@ unsigned long nfs_block_bits(unsigned long bsize, unsigned char *nrbitsp) /* * Calculate the number of 512byte blocks used. */ -static inline unsigned long nfs_calc_block_size(u64 tsize) +static inline blkcnt_t nfs_calc_block_size(u64 tsize) { - loff_t used = (tsize + 511) >> 9; + blkcnt_t used = (tsize + 511) >> 9; return (used > ULONG_MAX) ? ULONG_MAX : used; } diff --git a/fs/nfs/mount_clnt.c b/fs/nfs/mount_clnt.c index ca5a266a3140..8afd9f7e7a97 100644 --- a/fs/nfs/mount_clnt.c +++ b/fs/nfs/mount_clnt.c @@ -1,7 +1,5 @@ /* - * linux/fs/nfs/mount_clnt.c - * - * MOUNT client to support NFSroot. + * In-kernel MOUNT protocol client * * Copyright (C) 1997, Olaf Kirch <okir@monad.swb.de> */ @@ -18,33 +16,31 @@ #include <linux/nfs_fs.h> #ifdef RPC_DEBUG -# define NFSDBG_FACILITY NFSDBG_ROOT +# define NFSDBG_FACILITY NFSDBG_MOUNT #endif -/* -#define MOUNT_PROGRAM 100005 -#define MOUNT_VERSION 1 -#define MOUNT_MNT 1 -#define MOUNT_UMNT 3 - */ - -static struct rpc_clnt * mnt_create(char *, struct sockaddr_in *, - int, int); static struct rpc_program mnt_program; struct mnt_fhstatus { - unsigned int status; - struct nfs_fh * fh; + u32 status; + struct nfs_fh *fh; }; -/* - * Obtain an NFS file handle for the given host and path +/** + * nfs_mount - Obtain an NFS file handle for the given host and path + * @addr: pointer to server's address + * @len: size of server's address + * @hostname: name of server host, or NULL + * @path: pointer to string containing export path to mount + * @version: mount version to use for this request + * @protocol: transport protocol to use for thie request + * @fh: pointer to location to place returned file handle + * + * Uses default timeout parameters specified by underlying transport. */ -int -nfsroot_mount(struct sockaddr_in *addr, char *path, struct nfs_fh *fh, - int version, int protocol) +int nfs_mount(struct sockaddr *addr, size_t len, char *hostname, char *path, + int version, int protocol, struct nfs_fh *fh) { - struct rpc_clnt *mnt_clnt; struct mnt_fhstatus result = { .fh = fh }; @@ -52,16 +48,25 @@ nfsroot_mount(struct sockaddr_in *addr, char *path, struct nfs_fh *fh, .rpc_argp = path, .rpc_resp = &result, }; - char hostname[32]; + struct rpc_create_args args = { + .protocol = protocol, + .address = addr, + .addrsize = len, + .servername = hostname, + .program = &mnt_program, + .version = version, + .authflavor = RPC_AUTH_UNIX, + .flags = RPC_CLNT_CREATE_INTR, + }; + struct rpc_clnt *mnt_clnt; int status; - dprintk("NFS: nfs_mount(%08x:%s)\n", - (unsigned)ntohl(addr->sin_addr.s_addr), path); + dprintk("NFS: sending MNT request for %s:%s\n", + (hostname ? hostname : "server"), path); - sprintf(hostname, "%u.%u.%u.%u", NIPQUAD(addr->sin_addr.s_addr)); - mnt_clnt = mnt_create(hostname, addr, version, protocol); + mnt_clnt = rpc_create(&args); if (IS_ERR(mnt_clnt)) - return PTR_ERR(mnt_clnt); + goto out_clnt_err; if (version == NFS_MNT3_VERSION) msg.rpc_proc = &mnt_clnt->cl_procinfo[MOUNTPROC3_MNT]; @@ -69,33 +74,39 @@ nfsroot_mount(struct sockaddr_in *addr, char *path, struct nfs_fh *fh, msg.rpc_proc = &mnt_clnt->cl_procinfo[MNTPROC_MNT]; status = rpc_call_sync(mnt_clnt, &msg, 0); - return status < 0? status : (result.status? -EACCES : 0); -} + rpc_shutdown_client(mnt_clnt); -static struct rpc_clnt * -mnt_create(char *hostname, struct sockaddr_in *srvaddr, int version, - int protocol) -{ - struct rpc_create_args args = { - .protocol = protocol, - .address = (struct sockaddr *)srvaddr, - .addrsize = sizeof(*srvaddr), - .servername = hostname, - .program = &mnt_program, - .version = version, - .authflavor = RPC_AUTH_UNIX, - .flags = (RPC_CLNT_CREATE_ONESHOT | - RPC_CLNT_CREATE_INTR), - }; + if (status < 0) + goto out_call_err; + if (result.status != 0) + goto out_mnt_err; + + dprintk("NFS: MNT request succeeded\n"); + status = 0; + +out: + return status; + +out_clnt_err: + status = PTR_ERR(mnt_clnt); + dprintk("NFS: failed to create RPC client, status=%d\n", status); + goto out; + +out_call_err: + dprintk("NFS: failed to start MNT request, status=%d\n", status); + goto out; - return rpc_create(&args); +out_mnt_err: + dprintk("NFS: MNT server returned result %d\n", result.status); + status = -EACCES; + goto out; } /* * XDR encode/decode functions for MOUNT */ -static int -xdr_encode_dirpath(struct rpc_rqst *req, __be32 *p, const char *path) +static int xdr_encode_dirpath(struct rpc_rqst *req, __be32 *p, + const char *path) { p = xdr_encode_string(p, path); @@ -103,8 +114,8 @@ xdr_encode_dirpath(struct rpc_rqst *req, __be32 *p, const char *path) return 0; } -static int -xdr_decode_fhstatus(struct rpc_rqst *req, __be32 *p, struct mnt_fhstatus *res) +static int xdr_decode_fhstatus(struct rpc_rqst *req, __be32 *p, + struct mnt_fhstatus *res) { struct nfs_fh *fh = res->fh; @@ -115,8 +126,8 @@ xdr_decode_fhstatus(struct rpc_rqst *req, __be32 *p, struct mnt_fhstatus *res) return 0; } -static int -xdr_decode_fhstatus3(struct rpc_rqst *req, __be32 *p, struct mnt_fhstatus *res) +static int xdr_decode_fhstatus3(struct rpc_rqst *req, __be32 *p, + struct mnt_fhstatus *res) { struct nfs_fh *fh = res->fh; @@ -135,53 +146,53 @@ xdr_decode_fhstatus3(struct rpc_rqst *req, __be32 *p, struct mnt_fhstatus *res) #define MNT_fhstatus_sz (1 + 8) #define MNT_fhstatus3_sz (1 + 16) -static struct rpc_procinfo mnt_procedures[] = { -[MNTPROC_MNT] = { - .p_proc = MNTPROC_MNT, - .p_encode = (kxdrproc_t) xdr_encode_dirpath, - .p_decode = (kxdrproc_t) xdr_decode_fhstatus, - .p_arglen = MNT_dirpath_sz, - .p_replen = MNT_fhstatus_sz, - .p_statidx = MNTPROC_MNT, - .p_name = "MOUNT", +static struct rpc_procinfo mnt_procedures[] = { + [MNTPROC_MNT] = { + .p_proc = MNTPROC_MNT, + .p_encode = (kxdrproc_t) xdr_encode_dirpath, + .p_decode = (kxdrproc_t) xdr_decode_fhstatus, + .p_arglen = MNT_dirpath_sz, + .p_replen = MNT_fhstatus_sz, + .p_statidx = MNTPROC_MNT, + .p_name = "MOUNT", }, }; static struct rpc_procinfo mnt3_procedures[] = { -[MOUNTPROC3_MNT] = { - .p_proc = MOUNTPROC3_MNT, - .p_encode = (kxdrproc_t) xdr_encode_dirpath, - .p_decode = (kxdrproc_t) xdr_decode_fhstatus3, - .p_arglen = MNT_dirpath_sz, - .p_replen = MNT_fhstatus3_sz, - .p_statidx = MOUNTPROC3_MNT, - .p_name = "MOUNT", + [MOUNTPROC3_MNT] = { + .p_proc = MOUNTPROC3_MNT, + .p_encode = (kxdrproc_t) xdr_encode_dirpath, + .p_decode = (kxdrproc_t) xdr_decode_fhstatus3, + .p_arglen = MNT_dirpath_sz, + .p_replen = MNT_fhstatus3_sz, + .p_statidx = MOUNTPROC3_MNT, + .p_name = "MOUNT", }, }; -static struct rpc_version mnt_version1 = { - .number = 1, - .nrprocs = 2, - .procs = mnt_procedures +static struct rpc_version mnt_version1 = { + .number = 1, + .nrprocs = 2, + .procs = mnt_procedures, }; -static struct rpc_version mnt_version3 = { - .number = 3, - .nrprocs = 2, - .procs = mnt3_procedures +static struct rpc_version mnt_version3 = { + .number = 3, + .nrprocs = 2, + .procs = mnt3_procedures, }; -static struct rpc_version * mnt_version[] = { +static struct rpc_version *mnt_version[] = { NULL, &mnt_version1, NULL, &mnt_version3, }; -static struct rpc_stat mnt_stats; +static struct rpc_stat mnt_stats; -static struct rpc_program mnt_program = { +static struct rpc_program mnt_program = { .name = "mount", .number = NFS_MNT_PROGRAM, .nrvers = ARRAY_SIZE(mnt_version), diff --git a/fs/nfs/nfs2xdr.c b/fs/nfs/nfs2xdr.c index cd3ca7b5d3db..7fcc78f2aa71 100644 --- a/fs/nfs/nfs2xdr.c +++ b/fs/nfs/nfs2xdr.c @@ -223,7 +223,7 @@ nfs_xdr_diropargs(struct rpc_rqst *req, __be32 *p, struct nfs_diropargs *args) static int nfs_xdr_readargs(struct rpc_rqst *req, __be32 *p, struct nfs_readargs *args) { - struct rpc_auth *auth = req->rq_task->tk_auth; + struct rpc_auth *auth = req->rq_task->tk_msg.rpc_cred->cr_auth; unsigned int replen; u32 offset = (u32)args->offset; u32 count = args->count; @@ -380,7 +380,7 @@ static int nfs_xdr_readdirargs(struct rpc_rqst *req, __be32 *p, struct nfs_readdirargs *args) { struct rpc_task *task = req->rq_task; - struct rpc_auth *auth = task->tk_auth; + struct rpc_auth *auth = task->tk_msg.rpc_cred->cr_auth; unsigned int replen; u32 count = args->count; @@ -541,7 +541,7 @@ nfs_xdr_diropres(struct rpc_rqst *req, __be32 *p, struct nfs_diropok *res) static int nfs_xdr_readlinkargs(struct rpc_rqst *req, __be32 *p, struct nfs_readlinkargs *args) { - struct rpc_auth *auth = req->rq_task->tk_auth; + struct rpc_auth *auth = req->rq_task->tk_msg.rpc_cred->cr_auth; unsigned int replen; p = xdr_encode_fhandle(p, args->fh); diff --git a/fs/nfs/nfs3proc.c b/fs/nfs/nfs3proc.c index 45268d6def2e..814d886b6aa4 100644 --- a/fs/nfs/nfs3proc.c +++ b/fs/nfs/nfs3proc.c @@ -335,9 +335,7 @@ again: * not sure this buys us anything (and I'd have * to revamp the NFSv3 XDR code) */ status = nfs3_proc_setattr(dentry, &fattr, sattr); - if (status == 0) - nfs_setattr_update_inode(dentry->d_inode, sattr); - nfs_refresh_inode(dentry->d_inode, &fattr); + nfs_post_op_update_inode(dentry->d_inode, &fattr); dprintk("NFS reply setattr (post-create): %d\n", status); } if (status != 0) diff --git a/fs/nfs/nfs3xdr.c b/fs/nfs/nfs3xdr.c index b51df8eb9f01..b4647a22f349 100644 --- a/fs/nfs/nfs3xdr.c +++ b/fs/nfs/nfs3xdr.c @@ -319,7 +319,7 @@ nfs3_xdr_accessargs(struct rpc_rqst *req, __be32 *p, struct nfs3_accessargs *arg static int nfs3_xdr_readargs(struct rpc_rqst *req, __be32 *p, struct nfs_readargs *args) { - struct rpc_auth *auth = req->rq_task->tk_auth; + struct rpc_auth *auth = req->rq_task->tk_msg.rpc_cred->cr_auth; unsigned int replen; u32 count = args->count; @@ -458,7 +458,7 @@ nfs3_xdr_linkargs(struct rpc_rqst *req, __be32 *p, struct nfs3_linkargs *args) static int nfs3_xdr_readdirargs(struct rpc_rqst *req, __be32 *p, struct nfs3_readdirargs *args) { - struct rpc_auth *auth = req->rq_task->tk_auth; + struct rpc_auth *auth = req->rq_task->tk_msg.rpc_cred->cr_auth; unsigned int replen; u32 count = args->count; @@ -643,7 +643,7 @@ static int nfs3_xdr_getaclargs(struct rpc_rqst *req, __be32 *p, struct nfs3_getaclargs *args) { - struct rpc_auth *auth = req->rq_task->tk_auth; + struct rpc_auth *auth = req->rq_task->tk_msg.rpc_cred->cr_auth; unsigned int replen; p = xdr_encode_fhandle(p, args->fh); @@ -773,7 +773,7 @@ nfs3_xdr_accessres(struct rpc_rqst *req, __be32 *p, struct nfs3_accessres *res) static int nfs3_xdr_readlinkargs(struct rpc_rqst *req, __be32 *p, struct nfs3_readlinkargs *args) { - struct rpc_auth *auth = req->rq_task->tk_auth; + struct rpc_auth *auth = req->rq_task->tk_msg.rpc_cred->cr_auth; unsigned int replen; p = xdr_encode_fhandle(p, args->fh); diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h index cf3a17eb5c09..6c028e734fe6 100644 --- a/fs/nfs/nfs4_fs.h +++ b/fs/nfs/nfs4_fs.h @@ -70,19 +70,26 @@ static inline void nfs_confirm_seqid(struct nfs_seqid_counter *seqid, int status seqid->flags |= NFS_SEQID_CONFIRMED; } +struct nfs_unique_id { + struct rb_node rb_node; + __u64 id; +}; + /* * NFS4 state_owners and lock_owners are simply labels for ordered * sequences of RPC calls. Their sole purpose is to provide once-only * semantics by allowing the server to identify replayed requests. */ struct nfs4_state_owner { - spinlock_t so_lock; - struct list_head so_list; /* per-clientid list of state_owners */ + struct nfs_unique_id so_owner_id; struct nfs_client *so_client; - u32 so_id; /* 32-bit identifier, unique */ - atomic_t so_count; + struct nfs_server *so_server; + struct rb_node so_client_node; struct rpc_cred *so_cred; /* Associated cred */ + + spinlock_t so_lock; + atomic_t so_count; struct list_head so_states; struct list_head so_delegations; struct nfs_seqid_counter so_seqid; @@ -108,7 +115,7 @@ struct nfs4_lock_state { #define NFS_LOCK_INITIALIZED 1 int ls_flags; struct nfs_seqid_counter ls_seqid; - u32 ls_id; + struct nfs_unique_id ls_id; nfs4_stateid ls_stateid; atomic_t ls_count; }; @@ -116,7 +123,10 @@ struct nfs4_lock_state { /* bits for nfs4_state->flags */ enum { LK_STATE_IN_USE, - NFS_DELEGATED_STATE, + NFS_DELEGATED_STATE, /* Current stateid is delegation */ + NFS_O_RDONLY_STATE, /* OPEN stateid has read-only state */ + NFS_O_WRONLY_STATE, /* OPEN stateid has write-only state */ + NFS_O_RDWR_STATE, /* OPEN stateid has read/write state */ }; struct nfs4_state { @@ -130,11 +140,14 @@ struct nfs4_state { unsigned long flags; /* Do we hold any locks? */ spinlock_t state_lock; /* Protects the lock_states list */ - nfs4_stateid stateid; + seqlock_t seqlock; /* Protects the stateid/open_stateid */ + nfs4_stateid stateid; /* Current stateid: may be delegation */ + nfs4_stateid open_stateid; /* OPEN stateid */ - unsigned int n_rdonly; - unsigned int n_wronly; - unsigned int n_rdwr; + /* The following 3 fields are protected by owner->so_lock */ + unsigned int n_rdonly; /* Number of read-only references */ + unsigned int n_wronly; /* Number of write-only references */ + unsigned int n_rdwr; /* Number of read/write references */ int state; /* State on the server (R,W, or RW) */ atomic_t count; }; @@ -165,7 +178,7 @@ extern int nfs4_proc_setclientid(struct nfs_client *, u32, unsigned short, struc extern int nfs4_proc_setclientid_confirm(struct nfs_client *, struct rpc_cred *); extern int nfs4_proc_async_renew(struct nfs_client *, struct rpc_cred *); extern int nfs4_proc_renew(struct nfs_client *, struct rpc_cred *); -extern int nfs4_do_close(struct inode *inode, struct nfs4_state *state); +extern int nfs4_do_close(struct path *path, struct nfs4_state *state); extern struct dentry *nfs4_atomic_open(struct inode *, struct dentry *, struct nameidata *); extern int nfs4_open_revalidate(struct inode *, struct dentry *, int, struct nameidata *); extern int nfs4_server_capabilities(struct nfs_server *server, struct nfs_fh *fhandle); @@ -189,14 +202,13 @@ extern void nfs4_renew_state(struct work_struct *); /* nfs4state.c */ struct rpc_cred *nfs4_get_renew_cred(struct nfs_client *clp); -extern u32 nfs4_alloc_lockowner_id(struct nfs_client *); extern struct nfs4_state_owner * nfs4_get_state_owner(struct nfs_server *, struct rpc_cred *); extern void nfs4_put_state_owner(struct nfs4_state_owner *); extern void nfs4_drop_state_owner(struct nfs4_state_owner *); extern struct nfs4_state * nfs4_get_open_state(struct inode *, struct nfs4_state_owner *); extern void nfs4_put_open_state(struct nfs4_state *); -extern void nfs4_close_state(struct nfs4_state *, mode_t); +extern void nfs4_close_state(struct path *, struct nfs4_state *, mode_t); extern void nfs4_state_set_mode_locked(struct nfs4_state *, mode_t); extern void nfs4_schedule_state_recovery(struct nfs_client *); extern void nfs4_put_lock_state(struct nfs4_lock_state *lsp); @@ -222,7 +234,7 @@ extern struct svc_version nfs4_callback_version1; #else -#define nfs4_close_state(a, b) do { } while (0) +#define nfs4_close_state(a, b, c) do { } while (0) #endif /* CONFIG_NFS_V4 */ #endif /* __LINUX_FS_NFS_NFS4_FS.H */ diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 648e0ac0f90e..fee2da856c95 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -65,6 +65,7 @@ static int nfs4_async_handle_error(struct rpc_task *, const struct nfs_server *) static int _nfs4_proc_access(struct inode *inode, struct nfs_access_entry *entry); static int nfs4_handle_exception(const struct nfs_server *server, int errorcode, struct nfs4_exception *exception); static int nfs4_wait_clnt_recover(struct rpc_clnt *clnt, struct nfs_client *clp); +static int _nfs4_do_access(struct inode *inode, struct rpc_cred *cred, int openflags); /* Prevent leaks of NFSv4 errors into userland */ int nfs4_map_errors(int err) @@ -214,27 +215,39 @@ static void update_changeattr(struct inode *dir, struct nfs4_change_info *cinfo) } struct nfs4_opendata { - atomic_t count; + struct kref kref; struct nfs_openargs o_arg; struct nfs_openres o_res; struct nfs_open_confirmargs c_arg; struct nfs_open_confirmres c_res; struct nfs_fattr f_attr; struct nfs_fattr dir_attr; - struct dentry *dentry; + struct path path; struct dentry *dir; struct nfs4_state_owner *owner; + struct nfs4_state *state; struct iattr attrs; unsigned long timestamp; + unsigned int rpc_done : 1; int rpc_status; int cancelled; }; -static struct nfs4_opendata *nfs4_opendata_alloc(struct dentry *dentry, + +static void nfs4_init_opendata_res(struct nfs4_opendata *p) +{ + p->o_res.f_attr = &p->f_attr; + p->o_res.dir_attr = &p->dir_attr; + p->o_res.server = p->o_arg.server; + nfs_fattr_init(&p->f_attr); + nfs_fattr_init(&p->dir_attr); +} + +static struct nfs4_opendata *nfs4_opendata_alloc(struct path *path, struct nfs4_state_owner *sp, int flags, const struct iattr *attrs) { - struct dentry *parent = dget_parent(dentry); + struct dentry *parent = dget_parent(path->dentry); struct inode *dir = parent->d_inode; struct nfs_server *server = NFS_SERVER(dir); struct nfs4_opendata *p; @@ -245,24 +258,19 @@ static struct nfs4_opendata *nfs4_opendata_alloc(struct dentry *dentry, p->o_arg.seqid = nfs_alloc_seqid(&sp->so_seqid); if (p->o_arg.seqid == NULL) goto err_free; - atomic_set(&p->count, 1); - p->dentry = dget(dentry); + p->path.mnt = mntget(path->mnt); + p->path.dentry = dget(path->dentry); p->dir = parent; p->owner = sp; atomic_inc(&sp->so_count); p->o_arg.fh = NFS_FH(dir); p->o_arg.open_flags = flags, p->o_arg.clientid = server->nfs_client->cl_clientid; - p->o_arg.id = sp->so_id; - p->o_arg.name = &dentry->d_name; + p->o_arg.id = sp->so_owner_id.id; + p->o_arg.name = &p->path.dentry->d_name; p->o_arg.server = server; p->o_arg.bitmask = server->attr_bitmask; p->o_arg.claim = NFS4_OPEN_CLAIM_NULL; - p->o_res.f_attr = &p->f_attr; - p->o_res.dir_attr = &p->dir_attr; - p->o_res.server = server; - nfs_fattr_init(&p->f_attr); - nfs_fattr_init(&p->dir_attr); if (flags & O_EXCL) { u32 *s = (u32 *) p->o_arg.u.verifier.data; s[0] = jiffies; @@ -274,6 +282,8 @@ static struct nfs4_opendata *nfs4_opendata_alloc(struct dentry *dentry, p->c_arg.fh = &p->o_res.fh; p->c_arg.stateid = &p->o_res.stateid; p->c_arg.seqid = p->o_arg.seqid; + nfs4_init_opendata_res(p); + kref_init(&p->kref); return p; err_free: kfree(p); @@ -282,27 +292,25 @@ err: return NULL; } -static void nfs4_opendata_free(struct nfs4_opendata *p) +static void nfs4_opendata_free(struct kref *kref) { - if (p != NULL && atomic_dec_and_test(&p->count)) { - nfs_free_seqid(p->o_arg.seqid); - nfs4_put_state_owner(p->owner); - dput(p->dir); - dput(p->dentry); - kfree(p); - } + struct nfs4_opendata *p = container_of(kref, + struct nfs4_opendata, kref); + + nfs_free_seqid(p->o_arg.seqid); + if (p->state != NULL) + nfs4_put_open_state(p->state); + nfs4_put_state_owner(p->owner); + dput(p->dir); + dput(p->path.dentry); + mntput(p->path.mnt); + kfree(p); } -/* Helper for asynchronous RPC calls */ -static int nfs4_call_async(struct rpc_clnt *clnt, - const struct rpc_call_ops *tk_ops, void *calldata) +static void nfs4_opendata_put(struct nfs4_opendata *p) { - struct rpc_task *task; - - if (!(task = rpc_new_task(clnt, RPC_TASK_ASYNC, tk_ops, calldata))) - return -ENOMEM; - rpc_execute(task); - return 0; + if (p != NULL) + kref_put(&p->kref, nfs4_opendata_free); } static int nfs4_wait_for_completion_rpc_task(struct rpc_task *task) @@ -316,7 +324,34 @@ static int nfs4_wait_for_completion_rpc_task(struct rpc_task *task) return ret; } -static inline void update_open_stateflags(struct nfs4_state *state, mode_t open_flags) +static int can_open_cached(struct nfs4_state *state, int mode) +{ + int ret = 0; + switch (mode & (FMODE_READ|FMODE_WRITE|O_EXCL)) { + case FMODE_READ: + ret |= test_bit(NFS_O_RDONLY_STATE, &state->flags) != 0; + ret |= test_bit(NFS_O_RDWR_STATE, &state->flags) != 0; + break; + case FMODE_WRITE: + ret |= test_bit(NFS_O_WRONLY_STATE, &state->flags) != 0; + ret |= test_bit(NFS_O_RDWR_STATE, &state->flags) != 0; + break; + case FMODE_READ|FMODE_WRITE: + ret |= test_bit(NFS_O_RDWR_STATE, &state->flags) != 0; + } + return ret; +} + +static int can_open_delegated(struct nfs_delegation *delegation, mode_t open_flags) +{ + if ((delegation->type & open_flags) != open_flags) + return 0; + if (delegation->flags & NFS_DELEGATION_NEED_RECLAIM) + return 0; + return 1; +} + +static void update_open_stateflags(struct nfs4_state *state, mode_t open_flags) { switch (open_flags) { case FMODE_WRITE: @@ -328,41 +363,176 @@ static inline void update_open_stateflags(struct nfs4_state *state, mode_t open_ case FMODE_READ|FMODE_WRITE: state->n_rdwr++; } + nfs4_state_set_mode_locked(state, state->state | open_flags); } -static void update_open_stateid(struct nfs4_state *state, nfs4_stateid *stateid, int open_flags) +static void nfs_set_open_stateid_locked(struct nfs4_state *state, nfs4_stateid *stateid, int open_flags) { - struct inode *inode = state->inode; + if (test_bit(NFS_DELEGATED_STATE, &state->flags) == 0) + memcpy(state->stateid.data, stateid->data, sizeof(state->stateid.data)); + memcpy(state->open_stateid.data, stateid->data, sizeof(state->open_stateid.data)); + switch (open_flags) { + case FMODE_READ: + set_bit(NFS_O_RDONLY_STATE, &state->flags); + break; + case FMODE_WRITE: + set_bit(NFS_O_WRONLY_STATE, &state->flags); + break; + case FMODE_READ|FMODE_WRITE: + set_bit(NFS_O_RDWR_STATE, &state->flags); + } +} + +static void nfs_set_open_stateid(struct nfs4_state *state, nfs4_stateid *stateid, int open_flags) +{ + write_seqlock(&state->seqlock); + nfs_set_open_stateid_locked(state, stateid, open_flags); + write_sequnlock(&state->seqlock); +} +static void update_open_stateid(struct nfs4_state *state, nfs4_stateid *open_stateid, nfs4_stateid *deleg_stateid, int open_flags) +{ open_flags &= (FMODE_READ|FMODE_WRITE); - /* Protect against nfs4_find_state_byowner() */ + /* + * Protect the call to nfs4_state_set_mode_locked and + * serialise the stateid update + */ + write_seqlock(&state->seqlock); + if (deleg_stateid != NULL) { + memcpy(state->stateid.data, deleg_stateid->data, sizeof(state->stateid.data)); + set_bit(NFS_DELEGATED_STATE, &state->flags); + } + if (open_stateid != NULL) + nfs_set_open_stateid_locked(state, open_stateid, open_flags); + write_sequnlock(&state->seqlock); spin_lock(&state->owner->so_lock); - spin_lock(&inode->i_lock); - memcpy(&state->stateid, stateid, sizeof(state->stateid)); update_open_stateflags(state, open_flags); - nfs4_state_set_mode_locked(state, state->state | open_flags); - spin_unlock(&inode->i_lock); spin_unlock(&state->owner->so_lock); } +static void nfs4_return_incompatible_delegation(struct inode *inode, mode_t open_flags) +{ + struct nfs_delegation *delegation; + + rcu_read_lock(); + delegation = rcu_dereference(NFS_I(inode)->delegation); + if (delegation == NULL || (delegation->type & open_flags) == open_flags) { + rcu_read_unlock(); + return; + } + rcu_read_unlock(); + nfs_inode_return_delegation(inode); +} + +static struct nfs4_state *nfs4_try_open_cached(struct nfs4_opendata *opendata) +{ + struct nfs4_state *state = opendata->state; + struct nfs_inode *nfsi = NFS_I(state->inode); + struct nfs_delegation *delegation; + int open_mode = opendata->o_arg.open_flags & (FMODE_READ|FMODE_WRITE|O_EXCL); + nfs4_stateid stateid; + int ret = -EAGAIN; + + rcu_read_lock(); + delegation = rcu_dereference(nfsi->delegation); + for (;;) { + if (can_open_cached(state, open_mode)) { + spin_lock(&state->owner->so_lock); + if (can_open_cached(state, open_mode)) { + update_open_stateflags(state, open_mode); + spin_unlock(&state->owner->so_lock); + rcu_read_unlock(); + goto out_return_state; + } + spin_unlock(&state->owner->so_lock); + } + if (delegation == NULL) + break; + if (!can_open_delegated(delegation, open_mode)) + break; + /* Save the delegation */ + memcpy(stateid.data, delegation->stateid.data, sizeof(stateid.data)); + rcu_read_unlock(); + lock_kernel(); + ret = _nfs4_do_access(state->inode, state->owner->so_cred, open_mode); + unlock_kernel(); + if (ret != 0) + goto out; + ret = -EAGAIN; + rcu_read_lock(); + delegation = rcu_dereference(nfsi->delegation); + /* If no delegation, try a cached open */ + if (delegation == NULL) + continue; + /* Is the delegation still valid? */ + if (memcmp(stateid.data, delegation->stateid.data, sizeof(stateid.data)) != 0) + continue; + rcu_read_unlock(); + update_open_stateid(state, NULL, &stateid, open_mode); + goto out_return_state; + } + rcu_read_unlock(); +out: + return ERR_PTR(ret); +out_return_state: + atomic_inc(&state->count); + return state; +} + static struct nfs4_state *nfs4_opendata_to_nfs4_state(struct nfs4_opendata *data) { struct inode *inode; struct nfs4_state *state = NULL; + struct nfs_delegation *delegation; + nfs4_stateid *deleg_stateid = NULL; + int ret; - if (!(data->f_attr.valid & NFS_ATTR_FATTR)) + if (!data->rpc_done) { + state = nfs4_try_open_cached(data); goto out; + } + + ret = -EAGAIN; + if (!(data->f_attr.valid & NFS_ATTR_FATTR)) + goto err; inode = nfs_fhget(data->dir->d_sb, &data->o_res.fh, &data->f_attr); + ret = PTR_ERR(inode); if (IS_ERR(inode)) - goto out; + goto err; + ret = -ENOMEM; state = nfs4_get_open_state(inode, data->owner); if (state == NULL) - goto put_inode; - update_open_stateid(state, &data->o_res.stateid, data->o_arg.open_flags); -put_inode: + goto err_put_inode; + if (data->o_res.delegation_type != 0) { + int delegation_flags = 0; + + rcu_read_lock(); + delegation = rcu_dereference(NFS_I(inode)->delegation); + if (delegation) + delegation_flags = delegation->flags; + rcu_read_unlock(); + if (!(delegation_flags & NFS_DELEGATION_NEED_RECLAIM)) + nfs_inode_set_delegation(state->inode, + data->owner->so_cred, + &data->o_res); + else + nfs_inode_reclaim_delegation(state->inode, + data->owner->so_cred, + &data->o_res); + } + rcu_read_lock(); + delegation = rcu_dereference(NFS_I(inode)->delegation); + if (delegation != NULL) + deleg_stateid = &delegation->stateid; + update_open_stateid(state, &data->o_res.stateid, deleg_stateid, data->o_arg.open_flags); + rcu_read_unlock(); iput(inode); out: return state; +err_put_inode: + iput(inode); +err: + return ERR_PTR(ret); } static struct nfs_open_context *nfs4_state_find_open_context(struct nfs4_state *state) @@ -382,79 +552,66 @@ static struct nfs_open_context *nfs4_state_find_open_context(struct nfs4_state * return ERR_PTR(-ENOENT); } -static int nfs4_open_recover_helper(struct nfs4_opendata *opendata, mode_t openflags, nfs4_stateid *stateid) +static int nfs4_open_recover_helper(struct nfs4_opendata *opendata, mode_t openflags, struct nfs4_state **res) { + struct nfs4_state *newstate; int ret; opendata->o_arg.open_flags = openflags; + memset(&opendata->o_res, 0, sizeof(opendata->o_res)); + memset(&opendata->c_res, 0, sizeof(opendata->c_res)); + nfs4_init_opendata_res(opendata); ret = _nfs4_proc_open(opendata); if (ret != 0) return ret; - memcpy(stateid->data, opendata->o_res.stateid.data, - sizeof(stateid->data)); + newstate = nfs4_opendata_to_nfs4_state(opendata); + if (IS_ERR(newstate)) + return PTR_ERR(newstate); + nfs4_close_state(&opendata->path, newstate, openflags); + *res = newstate; return 0; } static int nfs4_open_recover(struct nfs4_opendata *opendata, struct nfs4_state *state) { - nfs4_stateid stateid; struct nfs4_state *newstate; - int mode = 0; - int delegation = 0; int ret; /* memory barrier prior to reading state->n_* */ + clear_bit(NFS_DELEGATED_STATE, &state->flags); smp_rmb(); if (state->n_rdwr != 0) { - ret = nfs4_open_recover_helper(opendata, FMODE_READ|FMODE_WRITE, &stateid); + ret = nfs4_open_recover_helper(opendata, FMODE_READ|FMODE_WRITE, &newstate); if (ret != 0) return ret; - mode |= FMODE_READ|FMODE_WRITE; - if (opendata->o_res.delegation_type != 0) - delegation = opendata->o_res.delegation_type; - smp_rmb(); + if (newstate != state) + return -ESTALE; } if (state->n_wronly != 0) { - ret = nfs4_open_recover_helper(opendata, FMODE_WRITE, &stateid); + ret = nfs4_open_recover_helper(opendata, FMODE_WRITE, &newstate); if (ret != 0) return ret; - mode |= FMODE_WRITE; - if (opendata->o_res.delegation_type != 0) - delegation = opendata->o_res.delegation_type; - smp_rmb(); + if (newstate != state) + return -ESTALE; } if (state->n_rdonly != 0) { - ret = nfs4_open_recover_helper(opendata, FMODE_READ, &stateid); + ret = nfs4_open_recover_helper(opendata, FMODE_READ, &newstate); if (ret != 0) return ret; - mode |= FMODE_READ; + if (newstate != state) + return -ESTALE; } - clear_bit(NFS_DELEGATED_STATE, &state->flags); - if (mode == 0) - return 0; - if (opendata->o_res.delegation_type == 0) - opendata->o_res.delegation_type = delegation; - opendata->o_arg.open_flags |= mode; - newstate = nfs4_opendata_to_nfs4_state(opendata); - if (newstate != NULL) { - if (opendata->o_res.delegation_type != 0) { - struct nfs_inode *nfsi = NFS_I(newstate->inode); - int delegation_flags = 0; - if (nfsi->delegation) - delegation_flags = nfsi->delegation->flags; - if (!(delegation_flags & NFS_DELEGATION_NEED_RECLAIM)) - nfs_inode_set_delegation(newstate->inode, - opendata->owner->so_cred, - &opendata->o_res); - else - nfs_inode_reclaim_delegation(newstate->inode, - opendata->owner->so_cred, - &opendata->o_res); - } - nfs4_close_state(newstate, opendata->o_arg.open_flags); + /* + * We may have performed cached opens for all three recoveries. + * Check if we need to update the current stateid. + */ + if (test_bit(NFS_DELEGATED_STATE, &state->flags) == 0 && + memcmp(state->stateid.data, state->open_stateid.data, sizeof(state->stateid.data)) != 0) { + write_seqlock(&state->seqlock); + if (test_bit(NFS_DELEGATED_STATE, &state->flags) == 0) + memcpy(state->stateid.data, state->open_stateid.data, sizeof(state->stateid.data)); + write_sequnlock(&state->seqlock); } - if (newstate != state) - return -ESTALE; return 0; } @@ -462,41 +619,37 @@ static int nfs4_open_recover(struct nfs4_opendata *opendata, struct nfs4_state * * OPEN_RECLAIM: * reclaim state on the server after a reboot. */ -static int _nfs4_do_open_reclaim(struct nfs4_state_owner *sp, struct nfs4_state *state, struct dentry *dentry) +static int _nfs4_do_open_reclaim(struct nfs_open_context *ctx, struct nfs4_state *state) { - struct nfs_delegation *delegation = NFS_I(state->inode)->delegation; + struct nfs_delegation *delegation; struct nfs4_opendata *opendata; int delegation_type = 0; int status; - if (delegation != NULL) { - if (!(delegation->flags & NFS_DELEGATION_NEED_RECLAIM)) { - memcpy(&state->stateid, &delegation->stateid, - sizeof(state->stateid)); - set_bit(NFS_DELEGATED_STATE, &state->flags); - return 0; - } - delegation_type = delegation->type; - } - opendata = nfs4_opendata_alloc(dentry, sp, 0, NULL); + opendata = nfs4_opendata_alloc(&ctx->path, state->owner, 0, NULL); if (opendata == NULL) return -ENOMEM; opendata->o_arg.claim = NFS4_OPEN_CLAIM_PREVIOUS; opendata->o_arg.fh = NFS_FH(state->inode); nfs_copy_fh(&opendata->o_res.fh, opendata->o_arg.fh); + rcu_read_lock(); + delegation = rcu_dereference(NFS_I(state->inode)->delegation); + if (delegation != NULL && (delegation->flags & NFS_DELEGATION_NEED_RECLAIM) != 0) + delegation_type = delegation->flags; + rcu_read_unlock(); opendata->o_arg.u.delegation_type = delegation_type; status = nfs4_open_recover(opendata, state); - nfs4_opendata_free(opendata); + nfs4_opendata_put(opendata); return status; } -static int nfs4_do_open_reclaim(struct nfs4_state_owner *sp, struct nfs4_state *state, struct dentry *dentry) +static int nfs4_do_open_reclaim(struct nfs_open_context *ctx, struct nfs4_state *state) { struct nfs_server *server = NFS_SERVER(state->inode); struct nfs4_exception exception = { }; int err; do { - err = _nfs4_do_open_reclaim(sp, state, dentry); + err = _nfs4_do_open_reclaim(ctx, state); if (err != -NFS4ERR_DELAY) break; nfs4_handle_exception(server, err, &exception); @@ -512,37 +665,35 @@ static int nfs4_open_reclaim(struct nfs4_state_owner *sp, struct nfs4_state *sta ctx = nfs4_state_find_open_context(state); if (IS_ERR(ctx)) return PTR_ERR(ctx); - ret = nfs4_do_open_reclaim(sp, state, ctx->dentry); + ret = nfs4_do_open_reclaim(ctx, state); put_nfs_open_context(ctx); return ret; } -static int _nfs4_open_delegation_recall(struct dentry *dentry, struct nfs4_state *state) +static int _nfs4_open_delegation_recall(struct nfs_open_context *ctx, struct nfs4_state *state, const nfs4_stateid *stateid) { struct nfs4_state_owner *sp = state->owner; struct nfs4_opendata *opendata; int ret; - if (!test_bit(NFS_DELEGATED_STATE, &state->flags)) - return 0; - opendata = nfs4_opendata_alloc(dentry, sp, 0, NULL); + opendata = nfs4_opendata_alloc(&ctx->path, sp, 0, NULL); if (opendata == NULL) return -ENOMEM; opendata->o_arg.claim = NFS4_OPEN_CLAIM_DELEGATE_CUR; - memcpy(opendata->o_arg.u.delegation.data, state->stateid.data, + memcpy(opendata->o_arg.u.delegation.data, stateid->data, sizeof(opendata->o_arg.u.delegation.data)); ret = nfs4_open_recover(opendata, state); - nfs4_opendata_free(opendata); + nfs4_opendata_put(opendata); return ret; } -int nfs4_open_delegation_recall(struct dentry *dentry, struct nfs4_state *state) +int nfs4_open_delegation_recall(struct nfs_open_context *ctx, struct nfs4_state *state, const nfs4_stateid *stateid) { struct nfs4_exception exception = { }; - struct nfs_server *server = NFS_SERVER(dentry->d_inode); + struct nfs_server *server = NFS_SERVER(state->inode); int err; do { - err = _nfs4_open_delegation_recall(dentry, state); + err = _nfs4_open_delegation_recall(ctx, state, stateid); switch (err) { case 0: return err; @@ -582,9 +733,10 @@ static void nfs4_open_confirm_done(struct rpc_task *task, void *calldata) memcpy(data->o_res.stateid.data, data->c_res.stateid.data, sizeof(data->o_res.stateid.data)); renew_lease(data->o_res.server, data->timestamp); + data->rpc_done = 1; } - nfs_increment_open_seqid(data->rpc_status, data->c_arg.seqid); nfs_confirm_seqid(&data->owner->so_seqid, data->rpc_status); + nfs_increment_open_seqid(data->rpc_status, data->c_arg.seqid); } static void nfs4_open_confirm_release(void *calldata) @@ -596,14 +748,14 @@ static void nfs4_open_confirm_release(void *calldata) if (data->cancelled == 0) goto out_free; /* In case of error, no cleanup! */ - if (data->rpc_status != 0) + if (!data->rpc_done) goto out_free; nfs_confirm_seqid(&data->owner->so_seqid, 0); state = nfs4_opendata_to_nfs4_state(data); - if (state != NULL) - nfs4_close_state(state, data->o_arg.open_flags); + if (!IS_ERR(state)) + nfs4_close_state(&data->path, state, data->o_arg.open_flags); out_free: - nfs4_opendata_free(data); + nfs4_opendata_put(data); } static const struct rpc_call_ops nfs4_open_confirm_ops = { @@ -621,12 +773,9 @@ static int _nfs4_proc_open_confirm(struct nfs4_opendata *data) struct rpc_task *task; int status; - atomic_inc(&data->count); - /* - * If rpc_run_task() ends up calling ->rpc_release(), we - * want to ensure that it takes the 'error' code path. - */ - data->rpc_status = -ENOMEM; + kref_get(&data->kref); + data->rpc_done = 0; + data->rpc_status = 0; task = rpc_run_task(server->client, RPC_TASK_ASYNC, &nfs4_open_confirm_ops, data); if (IS_ERR(task)) return PTR_ERR(task); @@ -653,13 +802,35 @@ static void nfs4_open_prepare(struct rpc_task *task, void *calldata) if (nfs_wait_on_sequence(data->o_arg.seqid, task) != 0) return; + /* + * Check if we still need to send an OPEN call, or if we can use + * a delegation instead. + */ + if (data->state != NULL) { + struct nfs_delegation *delegation; + + if (can_open_cached(data->state, data->o_arg.open_flags & (FMODE_READ|FMODE_WRITE|O_EXCL))) + goto out_no_action; + rcu_read_lock(); + delegation = rcu_dereference(NFS_I(data->state->inode)->delegation); + if (delegation != NULL && + (delegation->flags & NFS_DELEGATION_NEED_RECLAIM) == 0) { + rcu_read_unlock(); + goto out_no_action; + } + rcu_read_unlock(); + } /* Update sequence id. */ - data->o_arg.id = sp->so_id; + data->o_arg.id = sp->so_owner_id.id; data->o_arg.clientid = sp->so_client->cl_clientid; if (data->o_arg.claim == NFS4_OPEN_CLAIM_PREVIOUS) msg.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_OPEN_NOATTR]; data->timestamp = jiffies; rpc_call_setup(task, &msg, 0); + return; +out_no_action: + task->tk_action = NULL; + } static void nfs4_open_done(struct rpc_task *task, void *calldata) @@ -683,8 +854,11 @@ static void nfs4_open_done(struct rpc_task *task, void *calldata) data->rpc_status = -ENOTDIR; } renew_lease(data->o_res.server, data->timestamp); + if (!(data->o_res.rflags & NFS4_OPEN_RESULT_CONFIRM)) + nfs_confirm_seqid(&data->owner->so_seqid, 0); } nfs_increment_open_seqid(data->rpc_status, data->o_arg.seqid); + data->rpc_done = 1; } static void nfs4_open_release(void *calldata) @@ -696,17 +870,17 @@ static void nfs4_open_release(void *calldata) if (data->cancelled == 0) goto out_free; /* In case of error, no cleanup! */ - if (data->rpc_status != 0) + if (data->rpc_status != 0 || !data->rpc_done) goto out_free; /* In case we need an open_confirm, no cleanup! */ if (data->o_res.rflags & NFS4_OPEN_RESULT_CONFIRM) goto out_free; nfs_confirm_seqid(&data->owner->so_seqid, 0); state = nfs4_opendata_to_nfs4_state(data); - if (state != NULL) - nfs4_close_state(state, data->o_arg.open_flags); + if (!IS_ERR(state)) + nfs4_close_state(&data->path, state, data->o_arg.open_flags); out_free: - nfs4_opendata_free(data); + nfs4_opendata_put(data); } static const struct rpc_call_ops nfs4_open_ops = { @@ -727,12 +901,10 @@ static int _nfs4_proc_open(struct nfs4_opendata *data) struct rpc_task *task; int status; - atomic_inc(&data->count); - /* - * If rpc_run_task() ends up calling ->rpc_release(), we - * want to ensure that it takes the 'error' code path. - */ - data->rpc_status = -ENOMEM; + kref_get(&data->kref); + data->rpc_done = 0; + data->rpc_status = 0; + data->cancelled = 0; task = rpc_run_task(server->client, RPC_TASK_ASYNC, &nfs4_open_ops, data); if (IS_ERR(task)) return PTR_ERR(task); @@ -743,7 +915,7 @@ static int _nfs4_proc_open(struct nfs4_opendata *data) } else status = data->rpc_status; rpc_put_task(task); - if (status != 0) + if (status != 0 || !data->rpc_done) return status; if (o_arg->open_flags & O_CREAT) { @@ -756,7 +928,6 @@ static int _nfs4_proc_open(struct nfs4_opendata *data) if (status != 0) return status; } - nfs_confirm_seqid(&data->owner->so_seqid, 0); if (!(o_res->f_attr->valid & NFS_ATTR_FATTR)) return server->nfs_client->rpc_ops->getattr(server, &o_res->fh, o_res->f_attr); return 0; @@ -772,6 +943,8 @@ static int _nfs4_do_access(struct inode *inode, struct rpc_cred *cred, int openf mask |= MAY_READ; if (openflags & FMODE_WRITE) mask |= MAY_WRITE; + if (openflags & FMODE_EXEC) + mask |= MAY_EXEC; status = nfs_access_get_cached(inode, cred, &cache); if (status == 0) goto out; @@ -811,43 +984,32 @@ static int nfs4_recover_expired_lease(struct nfs_server *server) * reclaim state on the server after a network partition. * Assumes caller holds the appropriate lock */ -static int _nfs4_open_expired(struct nfs4_state_owner *sp, struct nfs4_state *state, struct dentry *dentry) +static int _nfs4_open_expired(struct nfs_open_context *ctx, struct nfs4_state *state) { - struct inode *inode = state->inode; - struct nfs_delegation *delegation = NFS_I(inode)->delegation; struct nfs4_opendata *opendata; - int openflags = state->state & (FMODE_READ|FMODE_WRITE); int ret; - if (delegation != NULL && !(delegation->flags & NFS_DELEGATION_NEED_RECLAIM)) { - ret = _nfs4_do_access(inode, sp->so_cred, openflags); - if (ret < 0) - return ret; - memcpy(&state->stateid, &delegation->stateid, sizeof(state->stateid)); - set_bit(NFS_DELEGATED_STATE, &state->flags); - return 0; - } - opendata = nfs4_opendata_alloc(dentry, sp, openflags, NULL); + opendata = nfs4_opendata_alloc(&ctx->path, state->owner, 0, NULL); if (opendata == NULL) return -ENOMEM; ret = nfs4_open_recover(opendata, state); if (ret == -ESTALE) { /* Invalidate the state owner so we don't ever use it again */ - nfs4_drop_state_owner(sp); - d_drop(dentry); + nfs4_drop_state_owner(state->owner); + d_drop(ctx->path.dentry); } - nfs4_opendata_free(opendata); + nfs4_opendata_put(opendata); return ret; } -static inline int nfs4_do_open_expired(struct nfs4_state_owner *sp, struct nfs4_state *state, struct dentry *dentry) +static inline int nfs4_do_open_expired(struct nfs_open_context *ctx, struct nfs4_state *state) { - struct nfs_server *server = NFS_SERVER(dentry->d_inode); + struct nfs_server *server = NFS_SERVER(state->inode); struct nfs4_exception exception = { }; int err; do { - err = _nfs4_open_expired(sp, state, dentry); + err = _nfs4_open_expired(ctx, state); if (err == -NFS4ERR_DELAY) nfs4_handle_exception(server, err, &exception); } while (exception.retry); @@ -862,107 +1024,38 @@ static int nfs4_open_expired(struct nfs4_state_owner *sp, struct nfs4_state *sta ctx = nfs4_state_find_open_context(state); if (IS_ERR(ctx)) return PTR_ERR(ctx); - ret = nfs4_do_open_expired(sp, state, ctx->dentry); + ret = nfs4_do_open_expired(ctx, state); put_nfs_open_context(ctx); return ret; } /* - * Returns a referenced nfs4_state if there is an open delegation on the file + * on an EXCLUSIVE create, the server should send back a bitmask with FATTR4-* + * fields corresponding to attributes that were used to store the verifier. + * Make sure we clobber those fields in the later setattr call */ -static int _nfs4_open_delegated(struct inode *inode, int flags, struct rpc_cred *cred, struct nfs4_state **res) -{ - struct nfs_delegation *delegation; - struct nfs_server *server = NFS_SERVER(inode); - struct nfs_client *clp = server->nfs_client; - struct nfs_inode *nfsi = NFS_I(inode); - struct nfs4_state_owner *sp = NULL; - struct nfs4_state *state = NULL; - int open_flags = flags & (FMODE_READ|FMODE_WRITE); - int err; - - err = -ENOMEM; - if (!(sp = nfs4_get_state_owner(server, cred))) { - dprintk("%s: nfs4_get_state_owner failed!\n", __FUNCTION__); - return err; - } - err = nfs4_recover_expired_lease(server); - if (err != 0) - goto out_put_state_owner; - /* Protect against reboot recovery - NOTE ORDER! */ - down_read(&clp->cl_sem); - /* Protect against delegation recall */ - down_read(&nfsi->rwsem); - delegation = NFS_I(inode)->delegation; - err = -ENOENT; - if (delegation == NULL || (delegation->type & open_flags) != open_flags) - goto out_err; - err = -ENOMEM; - state = nfs4_get_open_state(inode, sp); - if (state == NULL) - goto out_err; - - err = -ENOENT; - if ((state->state & open_flags) == open_flags) { - spin_lock(&inode->i_lock); - update_open_stateflags(state, open_flags); - spin_unlock(&inode->i_lock); - goto out_ok; - } else if (state->state != 0) - goto out_put_open_state; - - lock_kernel(); - err = _nfs4_do_access(inode, cred, open_flags); - unlock_kernel(); - if (err != 0) - goto out_put_open_state; - set_bit(NFS_DELEGATED_STATE, &state->flags); - update_open_stateid(state, &delegation->stateid, open_flags); -out_ok: - nfs4_put_state_owner(sp); - up_read(&nfsi->rwsem); - up_read(&clp->cl_sem); - *res = state; - return 0; -out_put_open_state: - nfs4_put_open_state(state); -out_err: - up_read(&nfsi->rwsem); - up_read(&clp->cl_sem); - if (err != -EACCES) - nfs_inode_return_delegation(inode); -out_put_state_owner: - nfs4_put_state_owner(sp); - return err; -} - -static struct nfs4_state *nfs4_open_delegated(struct inode *inode, int flags, struct rpc_cred *cred) +static inline void nfs4_exclusive_attrset(struct nfs4_opendata *opendata, struct iattr *sattr) { - struct nfs4_exception exception = { }; - struct nfs4_state *res = ERR_PTR(-EIO); - int err; + if ((opendata->o_res.attrset[1] & FATTR4_WORD1_TIME_ACCESS) && + !(sattr->ia_valid & ATTR_ATIME_SET)) + sattr->ia_valid |= ATTR_ATIME; - do { - err = _nfs4_open_delegated(inode, flags, cred, &res); - if (err == 0) - break; - res = ERR_PTR(nfs4_handle_exception(NFS_SERVER(inode), - err, &exception)); - } while (exception.retry); - return res; + if ((opendata->o_res.attrset[1] & FATTR4_WORD1_TIME_MODIFY) && + !(sattr->ia_valid & ATTR_MTIME_SET)) + sattr->ia_valid |= ATTR_MTIME; } /* * Returns a referenced nfs4_state */ -static int _nfs4_do_open(struct inode *dir, struct dentry *dentry, int flags, struct iattr *sattr, struct rpc_cred *cred, struct nfs4_state **res) +static int _nfs4_do_open(struct inode *dir, struct path *path, int flags, struct iattr *sattr, struct rpc_cred *cred, struct nfs4_state **res) { struct nfs4_state_owner *sp; struct nfs4_state *state = NULL; struct nfs_server *server = NFS_SERVER(dir); struct nfs_client *clp = server->nfs_client; struct nfs4_opendata *opendata; - int status; + int status; /* Protect against reboot recovery conflicts */ status = -ENOMEM; @@ -973,29 +1066,35 @@ static int _nfs4_do_open(struct inode *dir, struct dentry *dentry, int flags, st status = nfs4_recover_expired_lease(server); if (status != 0) goto err_put_state_owner; + if (path->dentry->d_inode != NULL) + nfs4_return_incompatible_delegation(path->dentry->d_inode, flags & (FMODE_READ|FMODE_WRITE)); down_read(&clp->cl_sem); status = -ENOMEM; - opendata = nfs4_opendata_alloc(dentry, sp, flags, sattr); + opendata = nfs4_opendata_alloc(path, sp, flags, sattr); if (opendata == NULL) goto err_release_rwsem; + if (path->dentry->d_inode != NULL) + opendata->state = nfs4_get_open_state(path->dentry->d_inode, sp); + status = _nfs4_proc_open(opendata); if (status != 0) - goto err_opendata_free; + goto err_opendata_put; + + if (opendata->o_arg.open_flags & O_EXCL) + nfs4_exclusive_attrset(opendata, sattr); - status = -ENOMEM; state = nfs4_opendata_to_nfs4_state(opendata); - if (state == NULL) - goto err_opendata_free; - if (opendata->o_res.delegation_type != 0) - nfs_inode_set_delegation(state->inode, cred, &opendata->o_res); - nfs4_opendata_free(opendata); + status = PTR_ERR(state); + if (IS_ERR(state)) + goto err_opendata_put; + nfs4_opendata_put(opendata); nfs4_put_state_owner(sp); up_read(&clp->cl_sem); *res = state; return 0; -err_opendata_free: - nfs4_opendata_free(opendata); +err_opendata_put: + nfs4_opendata_put(opendata); err_release_rwsem: up_read(&clp->cl_sem); err_put_state_owner: @@ -1006,14 +1105,14 @@ out_err: } -static struct nfs4_state *nfs4_do_open(struct inode *dir, struct dentry *dentry, int flags, struct iattr *sattr, struct rpc_cred *cred) +static struct nfs4_state *nfs4_do_open(struct inode *dir, struct path *path, int flags, struct iattr *sattr, struct rpc_cred *cred) { struct nfs4_exception exception = { }; struct nfs4_state *res; int status; do { - status = _nfs4_do_open(dir, dentry, flags, sattr, cred, &res); + status = _nfs4_do_open(dir, path, flags, sattr, cred, &res); if (status == 0) break; /* NOTE: BAD_SEQID means the server and client disagree about the @@ -1028,7 +1127,9 @@ static struct nfs4_state *nfs4_do_open(struct inode *dir, struct dentry *dentry, * the user though... */ if (status == -NFS4ERR_BAD_SEQID) { - printk(KERN_WARNING "NFS: v4 server returned a bad sequence-id error!\n"); + printk(KERN_WARNING "NFS: v4 server %s " + " returned a bad sequence-id error!\n", + NFS_SERVER(dir)->nfs_client->cl_hostname); exception.retry = 1; continue; } @@ -1042,6 +1143,11 @@ static struct nfs4_state *nfs4_do_open(struct inode *dir, struct dentry *dentry, exception.retry = 1; continue; } + if (status == -EAGAIN) { + /* We must have found a delegation */ + exception.retry = 1; + continue; + } res = ERR_PTR(nfs4_handle_exception(NFS_SERVER(dir), status, &exception)); } while (exception.retry); @@ -1101,6 +1207,7 @@ static int nfs4_do_setattr(struct inode *inode, struct nfs_fattr *fattr, } struct nfs4_closedata { + struct path path; struct inode *inode; struct nfs4_state *state; struct nfs_closeargs arg; @@ -1117,6 +1224,8 @@ static void nfs4_free_closedata(void *data) nfs4_put_open_state(calldata->state); nfs_free_seqid(calldata->arg.seqid); nfs4_put_state_owner(sp); + dput(calldata->path.dentry); + mntput(calldata->path.mnt); kfree(calldata); } @@ -1134,8 +1243,7 @@ static void nfs4_close_done(struct rpc_task *task, void *data) nfs_increment_open_seqid(task->tk_status, calldata->arg.seqid); switch (task->tk_status) { case 0: - memcpy(&state->stateid, &calldata->res.stateid, - sizeof(state->stateid)); + nfs_set_open_stateid(state, &calldata->res.stateid, calldata->arg.open_flags); renew_lease(server, calldata->timestamp); break; case -NFS4ERR_STALE_STATEID: @@ -1160,26 +1268,30 @@ static void nfs4_close_prepare(struct rpc_task *task, void *data) .rpc_resp = &calldata->res, .rpc_cred = state->owner->so_cred, }; - int mode = 0, old_mode; + int clear_rd, clear_wr, clear_rdwr; + int mode; if (nfs_wait_on_sequence(calldata->arg.seqid, task) != 0) return; - /* Recalculate the new open mode in case someone reopened the file - * while we were waiting in line to be scheduled. - */ + + mode = FMODE_READ|FMODE_WRITE; + clear_rd = clear_wr = clear_rdwr = 0; spin_lock(&state->owner->so_lock); - spin_lock(&calldata->inode->i_lock); - mode = old_mode = state->state; + /* Calculate the change in open mode */ if (state->n_rdwr == 0) { - if (state->n_rdonly == 0) + if (state->n_rdonly == 0) { mode &= ~FMODE_READ; - if (state->n_wronly == 0) + clear_rd |= test_and_clear_bit(NFS_O_RDONLY_STATE, &state->flags); + clear_rdwr |= test_and_clear_bit(NFS_O_RDWR_STATE, &state->flags); + } + if (state->n_wronly == 0) { mode &= ~FMODE_WRITE; + clear_wr |= test_and_clear_bit(NFS_O_WRONLY_STATE, &state->flags); + clear_rdwr |= test_and_clear_bit(NFS_O_RDWR_STATE, &state->flags); + } } - nfs4_state_set_mode_locked(state, mode); - spin_unlock(&calldata->inode->i_lock); spin_unlock(&state->owner->so_lock); - if (mode == old_mode || test_bit(NFS_DELEGATED_STATE, &state->flags)) { + if (!clear_rd && !clear_wr && !clear_rdwr) { /* Note: exit _without_ calling nfs4_close_done */ task->tk_action = NULL; return; @@ -1209,19 +1321,21 @@ static const struct rpc_call_ops nfs4_close_ops = { * * NOTE: Caller must be holding the sp->so_owner semaphore! */ -int nfs4_do_close(struct inode *inode, struct nfs4_state *state) +int nfs4_do_close(struct path *path, struct nfs4_state *state) { - struct nfs_server *server = NFS_SERVER(inode); + struct nfs_server *server = NFS_SERVER(state->inode); struct nfs4_closedata *calldata; + struct nfs4_state_owner *sp = state->owner; + struct rpc_task *task; int status = -ENOMEM; calldata = kmalloc(sizeof(*calldata), GFP_KERNEL); if (calldata == NULL) goto out; - calldata->inode = inode; + calldata->inode = state->inode; calldata->state = state; - calldata->arg.fh = NFS_FH(inode); - calldata->arg.stateid = &state->stateid; + calldata->arg.fh = NFS_FH(state->inode); + calldata->arg.stateid = &state->open_stateid; /* Serialization for the sequence id */ calldata->arg.seqid = nfs_alloc_seqid(&state->owner->so_seqid); if (calldata->arg.seqid == NULL) @@ -1229,36 +1343,55 @@ int nfs4_do_close(struct inode *inode, struct nfs4_state *state) calldata->arg.bitmask = server->attr_bitmask; calldata->res.fattr = &calldata->fattr; calldata->res.server = server; + calldata->path.mnt = mntget(path->mnt); + calldata->path.dentry = dget(path->dentry); - status = nfs4_call_async(server->client, &nfs4_close_ops, calldata); - if (status == 0) - goto out; - - nfs_free_seqid(calldata->arg.seqid); + task = rpc_run_task(server->client, RPC_TASK_ASYNC, &nfs4_close_ops, calldata); + if (IS_ERR(task)) + return PTR_ERR(task); + rpc_put_task(task); + return 0; out_free_calldata: kfree(calldata); out: + nfs4_put_open_state(state); + nfs4_put_state_owner(sp); return status; } -static int nfs4_intent_set_file(struct nameidata *nd, struct dentry *dentry, struct nfs4_state *state) +static int nfs4_intent_set_file(struct nameidata *nd, struct path *path, struct nfs4_state *state) { struct file *filp; + int ret; - filp = lookup_instantiate_filp(nd, dentry, NULL); + /* If the open_intent is for execute, we have an extra check to make */ + if (nd->intent.open.flags & FMODE_EXEC) { + ret = _nfs4_do_access(state->inode, + state->owner->so_cred, + nd->intent.open.flags); + if (ret < 0) + goto out_close; + } + filp = lookup_instantiate_filp(nd, path->dentry, NULL); if (!IS_ERR(filp)) { struct nfs_open_context *ctx; ctx = (struct nfs_open_context *)filp->private_data; ctx->state = state; return 0; } - nfs4_close_state(state, nd->intent.open.flags); - return PTR_ERR(filp); + ret = PTR_ERR(filp); +out_close: + nfs4_close_state(path, state, nd->intent.open.flags); + return ret; } struct dentry * nfs4_atomic_open(struct inode *dir, struct dentry *dentry, struct nameidata *nd) { + struct path path = { + .mnt = nd->mnt, + .dentry = dentry, + }; struct iattr attr; struct rpc_cred *cred; struct nfs4_state *state; @@ -1277,7 +1410,7 @@ nfs4_atomic_open(struct inode *dir, struct dentry *dentry, struct nameidata *nd) cred = rpcauth_lookupcred(NFS_CLIENT(dir)->cl_auth, 0); if (IS_ERR(cred)) return (struct dentry *)cred; - state = nfs4_do_open(dir, dentry, nd->intent.open.flags, &attr, cred); + state = nfs4_do_open(dir, &path, nd->intent.open.flags, &attr, cred); put_rpccred(cred); if (IS_ERR(state)) { if (PTR_ERR(state) == -ENOENT) @@ -1287,22 +1420,24 @@ nfs4_atomic_open(struct inode *dir, struct dentry *dentry, struct nameidata *nd) res = d_add_unique(dentry, igrab(state->inode)); if (res != NULL) dentry = res; - nfs4_intent_set_file(nd, dentry, state); + nfs4_intent_set_file(nd, &path, state); return res; } int nfs4_open_revalidate(struct inode *dir, struct dentry *dentry, int openflags, struct nameidata *nd) { + struct path path = { + .mnt = nd->mnt, + .dentry = dentry, + }; struct rpc_cred *cred; struct nfs4_state *state; cred = rpcauth_lookupcred(NFS_CLIENT(dir)->cl_auth, 0); if (IS_ERR(cred)) return PTR_ERR(cred); - state = nfs4_open_delegated(dentry->d_inode, openflags, cred); - if (IS_ERR(state)) - state = nfs4_do_open(dir, dentry, openflags, NULL, cred); + state = nfs4_do_open(dir, &path, openflags, NULL, cred); put_rpccred(cred); if (IS_ERR(state)) { switch (PTR_ERR(state)) { @@ -1318,10 +1453,10 @@ nfs4_open_revalidate(struct inode *dir, struct dentry *dentry, int openflags, st } } if (state->inode == dentry->d_inode) { - nfs4_intent_set_file(nd, dentry, state); + nfs4_intent_set_file(nd, &path, state); return 1; } - nfs4_close_state(state, openflags); + nfs4_close_state(&path, state, openflags); out_drop: d_drop(dentry); return 0; @@ -1559,8 +1694,6 @@ static int _nfs4_proc_lookupfh(struct nfs_server *server, struct nfs_fh *dirfh, dprintk("NFS call lookupfh %s\n", name->name); status = rpc_call_sync(server->client, &msg, 0); dprintk("NFS reply lookupfh: %d\n", status); - if (status == -NFS4ERR_MOVED) - status = -EREMOTE; return status; } @@ -1571,10 +1704,13 @@ static int nfs4_proc_lookupfh(struct nfs_server *server, struct nfs_fh *dirfh, struct nfs4_exception exception = { }; int err; do { - err = nfs4_handle_exception(server, - _nfs4_proc_lookupfh(server, dirfh, name, - fhandle, fattr), - &exception); + err = _nfs4_proc_lookupfh(server, dirfh, name, fhandle, fattr); + /* FIXME: !!!! */ + if (err == -NFS4ERR_MOVED) { + err = -EREMOTE; + break; + } + err = nfs4_handle_exception(server, err, &exception); } while (exception.retry); return err; } @@ -1582,28 +1718,10 @@ static int nfs4_proc_lookupfh(struct nfs_server *server, struct nfs_fh *dirfh, static int _nfs4_proc_lookup(struct inode *dir, struct qstr *name, struct nfs_fh *fhandle, struct nfs_fattr *fattr) { - int status; - struct nfs_server *server = NFS_SERVER(dir); - struct nfs4_lookup_arg args = { - .bitmask = server->attr_bitmask, - .dir_fh = NFS_FH(dir), - .name = name, - }; - struct nfs4_lookup_res res = { - .server = server, - .fattr = fattr, - .fh = fhandle, - }; - struct rpc_message msg = { - .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_LOOKUP], - .rpc_argp = &args, - .rpc_resp = &res, - }; - - nfs_fattr_init(fattr); + int status; dprintk("NFS call lookup %s\n", name->name); - status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); + status = _nfs4_proc_lookupfh(NFS_SERVER(dir), NFS_FH(dir), name, fhandle, fattr); if (status == -NFS4ERR_MOVED) status = nfs4_get_referral(dir, name, fattr, fhandle); dprintk("NFS reply lookup: %d\n", status); @@ -1752,6 +1870,10 @@ static int nfs4_proc_create(struct inode *dir, struct dentry *dentry, struct iattr *sattr, int flags, struct nameidata *nd) { + struct path path = { + .mnt = nd->mnt, + .dentry = dentry, + }; struct nfs4_state *state; struct rpc_cred *cred; int status = 0; @@ -1761,7 +1883,7 @@ nfs4_proc_create(struct inode *dir, struct dentry *dentry, struct iattr *sattr, status = PTR_ERR(cred); goto out; } - state = nfs4_do_open(dir, dentry, flags, sattr, cred); + state = nfs4_do_open(dir, &path, flags, sattr, cred); put_rpccred(cred); if (IS_ERR(state)) { status = PTR_ERR(state); @@ -1773,11 +1895,12 @@ nfs4_proc_create(struct inode *dir, struct dentry *dentry, struct iattr *sattr, status = nfs4_do_setattr(state->inode, &fattr, sattr, state); if (status == 0) nfs_setattr_update_inode(state->inode, sattr); + nfs_post_op_update_inode(state->inode, &fattr); } - if (status == 0 && nd != NULL && (nd->flags & LOOKUP_OPEN)) - status = nfs4_intent_set_file(nd, dentry, state); + if (status == 0 && (nd->flags & LOOKUP_OPEN) != 0) + status = nfs4_intent_set_file(nd, &path, state); else - nfs4_close_state(state, flags); + nfs4_close_state(&path, state, flags); out: return status; } @@ -3008,7 +3131,7 @@ static int _nfs4_proc_getlk(struct nfs4_state *state, int cmd, struct file_lock if (status != 0) goto out; lsp = request->fl_u.nfs4_fl.owner; - arg.lock_owner.id = lsp->ls_id; + arg.lock_owner.id = lsp->ls_id.id; status = rpc_call_sync(server->client, &msg, 0); switch (status) { case 0: @@ -3152,6 +3275,11 @@ static struct rpc_task *nfs4_do_unlck(struct file_lock *fl, { struct nfs4_unlockdata *data; + /* Ensure this is an unlock - when canceling a lock, the + * canceled lock is passed in, and it won't be an unlock. + */ + fl->fl_type = F_UNLCK; + data = nfs4_alloc_unlockdata(fl, ctx, lsp, seqid); if (data == NULL) { nfs_free_seqid(seqid); @@ -3222,7 +3350,7 @@ static struct nfs4_lockdata *nfs4_alloc_lockdata(struct file_lock *fl, goto out_free; p->arg.lock_stateid = &lsp->ls_stateid; p->arg.lock_owner.clientid = server->nfs_client->cl_clientid; - p->arg.lock_owner.id = lsp->ls_id; + p->arg.lock_owner.id = lsp->ls_id.id; p->lsp = lsp; atomic_inc(&lsp->ls_count); p->ctx = get_nfs_open_context(ctx); @@ -3285,7 +3413,7 @@ static void nfs4_lock_done(struct rpc_task *task, void *calldata) memcpy(data->lsp->ls_stateid.data, data->res.stateid.data, sizeof(data->lsp->ls_stateid.data)); data->lsp->ls_flags |= NFS_LOCK_INITIALIZED; - renew_lease(NFS_SERVER(data->ctx->dentry->d_inode), data->timestamp); + renew_lease(NFS_SERVER(data->ctx->path.dentry->d_inode), data->timestamp); } nfs_increment_lock_seqid(data->rpc_status, data->arg.lock_seqid); out: diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c index 8ed79d5c54f9..e9662ba81d86 100644 --- a/fs/nfs/nfs4state.c +++ b/fs/nfs/nfs4state.c @@ -38,12 +38,14 @@ * subsequent patch. */ +#include <linux/kernel.h> #include <linux/slab.h> #include <linux/smp_lock.h> #include <linux/nfs_fs.h> #include <linux/nfs_idmap.h> #include <linux/kthread.h> #include <linux/module.h> +#include <linux/random.h> #include <linux/workqueue.h> #include <linux/bitops.h> @@ -69,33 +71,14 @@ static int nfs4_init_client(struct nfs_client *clp, struct rpc_cred *cred) return status; } -u32 -nfs4_alloc_lockowner_id(struct nfs_client *clp) -{ - return clp->cl_lockowner_id ++; -} - -static struct nfs4_state_owner * -nfs4_client_grab_unused(struct nfs_client *clp, struct rpc_cred *cred) -{ - struct nfs4_state_owner *sp = NULL; - - if (!list_empty(&clp->cl_unused)) { - sp = list_entry(clp->cl_unused.next, struct nfs4_state_owner, so_list); - atomic_inc(&sp->so_count); - sp->so_cred = cred; - list_move(&sp->so_list, &clp->cl_state_owners); - clp->cl_nunused--; - } - return sp; -} - struct rpc_cred *nfs4_get_renew_cred(struct nfs_client *clp) { struct nfs4_state_owner *sp; + struct rb_node *pos; struct rpc_cred *cred = NULL; - list_for_each_entry(sp, &clp->cl_state_owners, so_list) { + for (pos = rb_first(&clp->cl_state_owners); pos != NULL; pos = rb_next(pos)) { + sp = rb_entry(pos, struct nfs4_state_owner, so_client_node); if (list_empty(&sp->so_states)) continue; cred = get_rpccred(sp->so_cred); @@ -107,32 +90,146 @@ struct rpc_cred *nfs4_get_renew_cred(struct nfs_client *clp) static struct rpc_cred *nfs4_get_setclientid_cred(struct nfs_client *clp) { struct nfs4_state_owner *sp; + struct rb_node *pos; - if (!list_empty(&clp->cl_state_owners)) { - sp = list_entry(clp->cl_state_owners.next, - struct nfs4_state_owner, so_list); + pos = rb_first(&clp->cl_state_owners); + if (pos != NULL) { + sp = rb_entry(pos, struct nfs4_state_owner, so_client_node); return get_rpccred(sp->so_cred); } return NULL; } +static void nfs_alloc_unique_id(struct rb_root *root, struct nfs_unique_id *new, + __u64 minval, int maxbits) +{ + struct rb_node **p, *parent; + struct nfs_unique_id *pos; + __u64 mask = ~0ULL; + + if (maxbits < 64) + mask = (1ULL << maxbits) - 1ULL; + + /* Ensure distribution is more or less flat */ + get_random_bytes(&new->id, sizeof(new->id)); + new->id &= mask; + if (new->id < minval) + new->id += minval; +retry: + p = &root->rb_node; + parent = NULL; + + while (*p != NULL) { + parent = *p; + pos = rb_entry(parent, struct nfs_unique_id, rb_node); + + if (new->id < pos->id) + p = &(*p)->rb_left; + else if (new->id > pos->id) + p = &(*p)->rb_right; + else + goto id_exists; + } + rb_link_node(&new->rb_node, parent, p); + rb_insert_color(&new->rb_node, root); + return; +id_exists: + for (;;) { + new->id++; + if (new->id < minval || (new->id & mask) != new->id) { + new->id = minval; + break; + } + parent = rb_next(parent); + if (parent == NULL) + break; + pos = rb_entry(parent, struct nfs_unique_id, rb_node); + if (new->id < pos->id) + break; + } + goto retry; +} + +static void nfs_free_unique_id(struct rb_root *root, struct nfs_unique_id *id) +{ + rb_erase(&id->rb_node, root); +} + static struct nfs4_state_owner * -nfs4_find_state_owner(struct nfs_client *clp, struct rpc_cred *cred) +nfs4_find_state_owner(struct nfs_server *server, struct rpc_cred *cred) { + struct nfs_client *clp = server->nfs_client; + struct rb_node **p = &clp->cl_state_owners.rb_node, + *parent = NULL; struct nfs4_state_owner *sp, *res = NULL; - list_for_each_entry(sp, &clp->cl_state_owners, so_list) { - if (sp->so_cred != cred) + while (*p != NULL) { + parent = *p; + sp = rb_entry(parent, struct nfs4_state_owner, so_client_node); + + if (server < sp->so_server) { + p = &parent->rb_left; continue; - atomic_inc(&sp->so_count); - /* Move to the head of the list */ - list_move(&sp->so_list, &clp->cl_state_owners); - res = sp; - break; + } + if (server > sp->so_server) { + p = &parent->rb_right; + continue; + } + if (cred < sp->so_cred) + p = &parent->rb_left; + else if (cred > sp->so_cred) + p = &parent->rb_right; + else { + atomic_inc(&sp->so_count); + res = sp; + break; + } } return res; } +static struct nfs4_state_owner * +nfs4_insert_state_owner(struct nfs_client *clp, struct nfs4_state_owner *new) +{ + struct rb_node **p = &clp->cl_state_owners.rb_node, + *parent = NULL; + struct nfs4_state_owner *sp; + + while (*p != NULL) { + parent = *p; + sp = rb_entry(parent, struct nfs4_state_owner, so_client_node); + + if (new->so_server < sp->so_server) { + p = &parent->rb_left; + continue; + } + if (new->so_server > sp->so_server) { + p = &parent->rb_right; + continue; + } + if (new->so_cred < sp->so_cred) + p = &parent->rb_left; + else if (new->so_cred > sp->so_cred) + p = &parent->rb_right; + else { + atomic_inc(&sp->so_count); + return sp; + } + } + nfs_alloc_unique_id(&clp->cl_openowner_id, &new->so_owner_id, 1, 64); + rb_link_node(&new->so_client_node, parent, p); + rb_insert_color(&new->so_client_node, &clp->cl_state_owners); + return new; +} + +static void +nfs4_remove_state_owner(struct nfs_client *clp, struct nfs4_state_owner *sp) +{ + if (!RB_EMPTY_NODE(&sp->so_client_node)) + rb_erase(&sp->so_client_node, &clp->cl_state_owners); + nfs_free_unique_id(&clp->cl_openowner_id, &sp->so_owner_id); +} + /* * nfs4_alloc_state_owner(): this is called on the OPEN or CREATE path to * create a new state_owner. @@ -160,10 +257,14 @@ nfs4_alloc_state_owner(void) void nfs4_drop_state_owner(struct nfs4_state_owner *sp) { - struct nfs_client *clp = sp->so_client; - spin_lock(&clp->cl_lock); - list_del_init(&sp->so_list); - spin_unlock(&clp->cl_lock); + if (!RB_EMPTY_NODE(&sp->so_client_node)) { + struct nfs_client *clp = sp->so_client; + + spin_lock(&clp->cl_lock); + rb_erase(&sp->so_client_node, &clp->cl_state_owners); + RB_CLEAR_NODE(&sp->so_client_node); + spin_unlock(&clp->cl_lock); + } } /* @@ -175,26 +276,25 @@ struct nfs4_state_owner *nfs4_get_state_owner(struct nfs_server *server, struct struct nfs_client *clp = server->nfs_client; struct nfs4_state_owner *sp, *new; - get_rpccred(cred); - new = nfs4_alloc_state_owner(); spin_lock(&clp->cl_lock); - sp = nfs4_find_state_owner(clp, cred); - if (sp == NULL) - sp = nfs4_client_grab_unused(clp, cred); - if (sp == NULL && new != NULL) { - list_add(&new->so_list, &clp->cl_state_owners); - new->so_client = clp; - new->so_id = nfs4_alloc_lockowner_id(clp); - new->so_cred = cred; - sp = new; - new = NULL; - } + sp = nfs4_find_state_owner(server, cred); spin_unlock(&clp->cl_lock); - kfree(new); if (sp != NULL) return sp; - put_rpccred(cred); - return NULL; + new = nfs4_alloc_state_owner(); + if (new == NULL) + return NULL; + new->so_client = clp; + new->so_server = server; + new->so_cred = cred; + spin_lock(&clp->cl_lock); + sp = nfs4_insert_state_owner(clp, new); + spin_unlock(&clp->cl_lock); + if (sp == new) + get_rpccred(cred); + else + kfree(new); + return sp; } /* @@ -208,18 +308,7 @@ void nfs4_put_state_owner(struct nfs4_state_owner *sp) if (!atomic_dec_and_lock(&sp->so_count, &clp->cl_lock)) return; - if (clp->cl_nunused >= OPENOWNER_POOL_SIZE) - goto out_free; - if (list_empty(&sp->so_list)) - goto out_free; - list_move(&sp->so_list, &clp->cl_unused); - clp->cl_nunused++; - spin_unlock(&clp->cl_lock); - put_rpccred(cred); - cred = NULL; - return; -out_free: - list_del(&sp->so_list); + nfs4_remove_state_owner(clp, sp); spin_unlock(&clp->cl_lock); put_rpccred(cred); kfree(sp); @@ -236,6 +325,7 @@ nfs4_alloc_open_state(void) atomic_set(&state->count, 1); INIT_LIST_HEAD(&state->lock_states); spin_lock_init(&state->state_lock); + seqlock_init(&state->seqlock); return state; } @@ -263,13 +353,10 @@ __nfs4_find_state_byowner(struct inode *inode, struct nfs4_state_owner *owner) struct nfs4_state *state; list_for_each_entry(state, &nfsi->open_states, inode_states) { - /* Is this in the process of being freed? */ - if (state->state == 0) + if (state->owner != owner) continue; - if (state->owner == owner) { - atomic_inc(&state->count); + if (atomic_inc_not_zero(&state->count)) return state; - } } return NULL; } @@ -341,16 +428,15 @@ void nfs4_put_open_state(struct nfs4_state *state) /* * Close the current file. */ -void nfs4_close_state(struct nfs4_state *state, mode_t mode) +void nfs4_close_state(struct path *path, struct nfs4_state *state, mode_t mode) { - struct inode *inode = state->inode; struct nfs4_state_owner *owner = state->owner; - int oldstate, newstate = 0; + int call_close = 0; + int newstate; atomic_inc(&owner->so_count); /* Protect against nfs4_find_state() */ spin_lock(&owner->so_lock); - spin_lock(&inode->i_lock); switch (mode & (FMODE_READ | FMODE_WRITE)) { case FMODE_READ: state->n_rdonly--; @@ -361,24 +447,29 @@ void nfs4_close_state(struct nfs4_state *state, mode_t mode) case FMODE_READ|FMODE_WRITE: state->n_rdwr--; } - oldstate = newstate = state->state; + newstate = FMODE_READ|FMODE_WRITE; if (state->n_rdwr == 0) { - if (state->n_rdonly == 0) + if (state->n_rdonly == 0) { newstate &= ~FMODE_READ; - if (state->n_wronly == 0) + call_close |= test_bit(NFS_O_RDONLY_STATE, &state->flags); + call_close |= test_bit(NFS_O_RDWR_STATE, &state->flags); + } + if (state->n_wronly == 0) { newstate &= ~FMODE_WRITE; + call_close |= test_bit(NFS_O_WRONLY_STATE, &state->flags); + call_close |= test_bit(NFS_O_RDWR_STATE, &state->flags); + } + if (newstate == 0) + clear_bit(NFS_DELEGATED_STATE, &state->flags); } - if (test_bit(NFS_DELEGATED_STATE, &state->flags)) { - nfs4_state_set_mode_locked(state, newstate); - oldstate = newstate; - } - spin_unlock(&inode->i_lock); + nfs4_state_set_mode_locked(state, newstate); spin_unlock(&owner->so_lock); - if (oldstate != newstate && nfs4_do_close(inode, state) == 0) - return; - nfs4_put_open_state(state); - nfs4_put_state_owner(owner); + if (!call_close) { + nfs4_put_open_state(state); + nfs4_put_state_owner(owner); + } else + nfs4_do_close(path, state); } /* @@ -415,12 +506,22 @@ static struct nfs4_lock_state *nfs4_alloc_lock_state(struct nfs4_state *state, f atomic_set(&lsp->ls_count, 1); lsp->ls_owner = fl_owner; spin_lock(&clp->cl_lock); - lsp->ls_id = nfs4_alloc_lockowner_id(clp); + nfs_alloc_unique_id(&clp->cl_lockowner_id, &lsp->ls_id, 1, 64); spin_unlock(&clp->cl_lock); INIT_LIST_HEAD(&lsp->ls_locks); return lsp; } +static void nfs4_free_lock_state(struct nfs4_lock_state *lsp) +{ + struct nfs_client *clp = lsp->ls_state->owner->so_client; + + spin_lock(&clp->cl_lock); + nfs_free_unique_id(&clp->cl_lockowner_id, &lsp->ls_id); + spin_unlock(&clp->cl_lock); + kfree(lsp); +} + /* * Return a compatible lock_state. If no initialized lock_state structure * exists, return an uninitialized one. @@ -450,7 +551,8 @@ static struct nfs4_lock_state *nfs4_get_lock_state(struct nfs4_state *state, fl_ return NULL; } spin_unlock(&state->state_lock); - kfree(new); + if (new != NULL) + nfs4_free_lock_state(new); return lsp; } @@ -471,7 +573,7 @@ void nfs4_put_lock_state(struct nfs4_lock_state *lsp) if (list_empty(&state->lock_states)) clear_bit(LK_STATE_IN_USE, &state->flags); spin_unlock(&state->state_lock); - kfree(lsp); + nfs4_free_lock_state(lsp); } static void nfs4_fl_copy_lock(struct file_lock *dst, struct file_lock *src) @@ -513,8 +615,12 @@ int nfs4_set_lock_state(struct nfs4_state *state, struct file_lock *fl) void nfs4_copy_stateid(nfs4_stateid *dst, struct nfs4_state *state, fl_owner_t fl_owner) { struct nfs4_lock_state *lsp; + int seq; - memcpy(dst, &state->stateid, sizeof(*dst)); + do { + seq = read_seqbegin(&state->seqlock); + memcpy(dst, &state->stateid, sizeof(*dst)); + } while (read_seqretry(&state->seqlock, seq)); if (test_bit(LK_STATE_IN_USE, &state->flags) == 0) return; @@ -557,12 +663,18 @@ void nfs_free_seqid(struct nfs_seqid *seqid) * failed with a seqid incrementing error - * see comments nfs_fs.h:seqid_mutating_error() */ -static inline void nfs_increment_seqid(int status, struct nfs_seqid *seqid) +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; + printk(KERN_WARNING "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: @@ -586,7 +698,7 @@ void nfs_increment_open_seqid(int status, struct nfs_seqid *seqid) struct nfs4_state_owner, so_seqid); nfs4_drop_state_owner(sp); } - return nfs_increment_seqid(status, seqid); + nfs_increment_seqid(status, seqid); } /* @@ -596,7 +708,7 @@ void nfs_increment_open_seqid(int status, struct nfs_seqid *seqid) */ void nfs_increment_lock_seqid(int status, struct nfs_seqid *seqid) { - return nfs_increment_seqid(status, seqid); + nfs_increment_seqid(status, seqid); } int nfs_wait_on_sequence(struct nfs_seqid *seqid, struct rpc_task *task) @@ -748,15 +860,21 @@ out_err: static void nfs4_state_mark_reclaim(struct nfs_client *clp) { struct nfs4_state_owner *sp; + struct rb_node *pos; struct nfs4_state *state; struct nfs4_lock_state *lock; /* Reset all sequence ids to zero */ - list_for_each_entry(sp, &clp->cl_state_owners, so_list) { + for (pos = rb_first(&clp->cl_state_owners); pos != NULL; pos = rb_next(pos)) { + sp = rb_entry(pos, struct nfs4_state_owner, so_client_node); sp->so_seqid.counter = 0; sp->so_seqid.flags = 0; spin_lock(&sp->so_lock); list_for_each_entry(state, &sp->so_states, open_states) { + 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); list_for_each_entry(lock, &state->lock_states, ls_locks) { lock->ls_seqid.counter = 0; lock->ls_seqid.flags = 0; @@ -771,6 +889,7 @@ static int reclaimer(void *ptr) { struct nfs_client *clp = ptr; struct nfs4_state_owner *sp; + struct rb_node *pos; struct nfs4_state_recovery_ops *ops; struct rpc_cred *cred; int status = 0; @@ -816,7 +935,8 @@ restart_loop: /* Mark all delegations for reclaim */ nfs_delegation_mark_reclaim(clp); /* Note: list is protected by exclusive lock on cl->cl_sem */ - list_for_each_entry(sp, &clp->cl_state_owners, so_list) { + for (pos = rb_first(&clp->cl_state_owners); pos != NULL; pos = rb_next(pos)) { + sp = rb_entry(pos, struct nfs4_state_owner, so_client_node); status = nfs4_reclaim_open_state(ops, sp); if (status < 0) { if (status == -NFS4ERR_NO_GRACE) { diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index 8003c91ccb9a..c08738441f73 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -68,9 +68,10 @@ static int nfs4_stat_to_errno(int); #endif /* lock,open owner id: - * we currently use size 1 (u32) out of (NFS4_OPAQUE_LIMIT >> 2) + * we currently use size 2 (u64) out of (NFS4_OPAQUE_LIMIT >> 2) */ -#define owner_id_maxsz (1 + 1) +#define open_owner_id_maxsz (1 + 4) +#define lock_owner_id_maxsz (1 + 4) #define compound_encode_hdr_maxsz (3 + (NFS4_MAXTAGLEN >> 2)) #define compound_decode_hdr_maxsz (3 + (NFS4_MAXTAGLEN >> 2)) #define op_encode_hdr_maxsz (1) @@ -87,9 +88,11 @@ static int nfs4_stat_to_errno(int); #define encode_getattr_maxsz (op_encode_hdr_maxsz + nfs4_fattr_bitmap_maxsz) #define nfs4_name_maxsz (1 + ((3 + NFS4_MAXNAMLEN) >> 2)) #define nfs4_path_maxsz (1 + ((3 + NFS4_MAXPATHLEN) >> 2)) +#define nfs4_owner_maxsz (1 + XDR_QUADLEN(IDMAP_NAMESZ)) +#define nfs4_group_maxsz (1 + XDR_QUADLEN(IDMAP_NAMESZ)) /* This is based on getfattr, which uses the most attributes: */ #define nfs4_fattr_value_maxsz (1 + (1 + 2 + 2 + 4 + 2 + 1 + 1 + 2 + 2 + \ - 3 + 3 + 3 + 2 * nfs4_name_maxsz)) + 3 + 3 + 3 + nfs4_owner_maxsz + nfs4_group_maxsz)) #define nfs4_fattr_maxsz (nfs4_fattr_bitmap_maxsz + \ nfs4_fattr_value_maxsz) #define decode_getattr_maxsz (op_decode_hdr_maxsz + nfs4_fattr_maxsz) @@ -116,8 +119,27 @@ static int nfs4_stat_to_errno(int); 3 + (NFS4_VERIFIER_SIZE >> 2)) #define decode_setclientid_confirm_maxsz \ (op_decode_hdr_maxsz) -#define encode_lookup_maxsz (op_encode_hdr_maxsz + \ - 1 + ((3 + NFS4_FHSIZE) >> 2)) +#define encode_lookup_maxsz (op_encode_hdr_maxsz + nfs4_name_maxsz) +#define decode_lookup_maxsz (op_decode_hdr_maxsz) +#define encode_share_access_maxsz \ + (2) +#define encode_createmode_maxsz (1 + nfs4_fattr_maxsz) +#define encode_opentype_maxsz (1 + encode_createmode_maxsz) +#define encode_claim_null_maxsz (1 + nfs4_name_maxsz) +#define encode_open_maxsz (op_encode_hdr_maxsz + \ + 2 + encode_share_access_maxsz + 2 + \ + open_owner_id_maxsz + \ + encode_opentype_maxsz + \ + encode_claim_null_maxsz) +#define decode_ace_maxsz (3 + nfs4_owner_maxsz) +#define decode_delegation_maxsz (1 + XDR_QUADLEN(NFS4_STATEID_SIZE) + 1 + \ + decode_ace_maxsz) +#define decode_change_info_maxsz (5) +#define decode_open_maxsz (op_decode_hdr_maxsz + \ + XDR_QUADLEN(NFS4_STATEID_SIZE) + \ + decode_change_info_maxsz + 1 + \ + nfs4_fattr_bitmap_maxsz + \ + decode_delegation_maxsz) #define encode_remove_maxsz (op_encode_hdr_maxsz + \ nfs4_name_maxsz) #define encode_rename_maxsz (op_encode_hdr_maxsz + \ @@ -134,9 +156,15 @@ static int nfs4_stat_to_errno(int); #define encode_create_maxsz (op_encode_hdr_maxsz + \ 2 + nfs4_name_maxsz + \ nfs4_fattr_maxsz) -#define decode_create_maxsz (op_decode_hdr_maxsz + 8) +#define decode_create_maxsz (op_decode_hdr_maxsz + \ + decode_change_info_maxsz + \ + nfs4_fattr_bitmap_maxsz) #define encode_delegreturn_maxsz (op_encode_hdr_maxsz + 4) #define decode_delegreturn_maxsz (op_decode_hdr_maxsz) +#define encode_fs_locations_maxsz \ + (encode_getattr_maxsz) +#define decode_fs_locations_maxsz \ + (0) #define NFS4_enc_compound_sz (1024) /* XXX: large enough? */ #define NFS4_dec_compound_sz (1024) /* XXX: large enough? */ #define NFS4_enc_read_sz (compound_encode_hdr_maxsz + \ @@ -174,16 +202,21 @@ static int nfs4_stat_to_errno(int); op_decode_hdr_maxsz + 2 + \ decode_getattr_maxsz) #define NFS4_enc_open_sz (compound_encode_hdr_maxsz + \ - encode_putfh_maxsz + \ - op_encode_hdr_maxsz + \ - 13 + 3 + 2 + 64 + \ - encode_getattr_maxsz + \ - encode_getfh_maxsz) + encode_putfh_maxsz + \ + encode_savefh_maxsz + \ + encode_open_maxsz + \ + encode_getfh_maxsz + \ + encode_getattr_maxsz + \ + encode_restorefh_maxsz + \ + encode_getattr_maxsz) #define NFS4_dec_open_sz (compound_decode_hdr_maxsz + \ - decode_putfh_maxsz + \ - op_decode_hdr_maxsz + 4 + 5 + 2 + 3 + \ - decode_getattr_maxsz + \ - decode_getfh_maxsz) + decode_putfh_maxsz + \ + decode_savefh_maxsz + \ + decode_open_maxsz + \ + decode_getfh_maxsz + \ + decode_getattr_maxsz + \ + decode_restorefh_maxsz + \ + decode_getattr_maxsz) #define NFS4_enc_open_confirm_sz \ (compound_encode_hdr_maxsz + \ encode_putfh_maxsz + \ @@ -193,12 +226,12 @@ static int nfs4_stat_to_errno(int); op_decode_hdr_maxsz + 4) #define NFS4_enc_open_noattr_sz (compound_encode_hdr_maxsz + \ encode_putfh_maxsz + \ - op_encode_hdr_maxsz + \ - 11) + encode_open_maxsz + \ + encode_getattr_maxsz) #define NFS4_dec_open_noattr_sz (compound_decode_hdr_maxsz + \ decode_putfh_maxsz + \ - op_decode_hdr_maxsz + \ - 4 + 5 + 2 + 3) + decode_open_maxsz + \ + decode_getattr_maxsz) #define NFS4_enc_open_downgrade_sz \ (compound_encode_hdr_maxsz + \ encode_putfh_maxsz + \ @@ -256,19 +289,19 @@ static int nfs4_stat_to_errno(int); op_encode_hdr_maxsz + \ 1 + 1 + 2 + 2 + \ 1 + 4 + 1 + 2 + \ - owner_id_maxsz) + lock_owner_id_maxsz) #define NFS4_dec_lock_sz (compound_decode_hdr_maxsz + \ decode_putfh_maxsz + \ decode_getattr_maxsz + \ op_decode_hdr_maxsz + \ 2 + 2 + 1 + 2 + \ - owner_id_maxsz) + lock_owner_id_maxsz) #define NFS4_enc_lockt_sz (compound_encode_hdr_maxsz + \ encode_putfh_maxsz + \ encode_getattr_maxsz + \ op_encode_hdr_maxsz + \ 1 + 2 + 2 + 2 + \ - owner_id_maxsz) + lock_owner_id_maxsz) #define NFS4_dec_lockt_sz (NFS4_dec_lock_sz) #define NFS4_enc_locku_sz (compound_encode_hdr_maxsz + \ encode_putfh_maxsz + \ @@ -298,7 +331,7 @@ static int nfs4_stat_to_errno(int); encode_getfh_maxsz) #define NFS4_dec_lookup_sz (compound_decode_hdr_maxsz + \ decode_putfh_maxsz + \ - op_decode_hdr_maxsz + \ + decode_lookup_maxsz + \ decode_getattr_maxsz + \ decode_getfh_maxsz) #define NFS4_enc_lookup_root_sz (compound_encode_hdr_maxsz + \ @@ -417,12 +450,13 @@ static int nfs4_stat_to_errno(int); #define NFS4_enc_fs_locations_sz \ (compound_encode_hdr_maxsz + \ encode_putfh_maxsz + \ - encode_getattr_maxsz) + encode_lookup_maxsz + \ + encode_fs_locations_maxsz) #define NFS4_dec_fs_locations_sz \ (compound_decode_hdr_maxsz + \ decode_putfh_maxsz + \ - op_decode_hdr_maxsz + \ - nfs4_fattr_bitmap_maxsz) + decode_lookup_maxsz + \ + decode_fs_locations_maxsz) static struct { unsigned int mode; @@ -793,13 +827,14 @@ static int encode_lock(struct xdr_stream *xdr, const struct nfs_lock_args *args) WRITE64(nfs4_lock_length(args->fl)); WRITE32(args->new_lock_owner); if (args->new_lock_owner){ - RESERVE_SPACE(4+NFS4_STATEID_SIZE+20); + RESERVE_SPACE(4+NFS4_STATEID_SIZE+32); WRITE32(args->open_seqid->sequence->counter); WRITEMEM(args->open_stateid->data, NFS4_STATEID_SIZE); WRITE32(args->lock_seqid->sequence->counter); WRITE64(args->lock_owner.clientid); - WRITE32(4); - WRITE32(args->lock_owner.id); + WRITE32(16); + WRITEMEM("lock id:", 8); + WRITE64(args->lock_owner.id); } else { RESERVE_SPACE(NFS4_STATEID_SIZE+4); @@ -814,14 +849,15 @@ static int encode_lockt(struct xdr_stream *xdr, const struct nfs_lockt_args *arg { __be32 *p; - RESERVE_SPACE(40); + RESERVE_SPACE(52); WRITE32(OP_LOCKT); WRITE32(nfs4_lock_type(args->fl, 0)); WRITE64(args->fl->fl_start); WRITE64(nfs4_lock_length(args->fl)); WRITE64(args->lock_owner.clientid); - WRITE32(4); - WRITE32(args->lock_owner.id); + WRITE32(16); + WRITEMEM("lock id:", 8); + WRITE64(args->lock_owner.id); return 0; } @@ -886,10 +922,11 @@ static inline void encode_openhdr(struct xdr_stream *xdr, const struct nfs_opena WRITE32(OP_OPEN); WRITE32(arg->seqid->sequence->counter); encode_share_access(xdr, arg->open_flags); - RESERVE_SPACE(16); + RESERVE_SPACE(28); WRITE64(arg->clientid); - WRITE32(4); - WRITE32(arg->id); + WRITE32(16); + WRITEMEM("open id:", 8); + WRITE64(arg->id); } static inline void encode_createmode(struct xdr_stream *xdr, const struct nfs_openargs *arg) @@ -1071,7 +1108,7 @@ static int encode_read(struct xdr_stream *xdr, const struct nfs_readargs *args) static int encode_readdir(struct xdr_stream *xdr, const struct nfs4_readdir_arg *readdir, struct rpc_rqst *req) { - struct rpc_auth *auth = req->rq_task->tk_auth; + struct rpc_auth *auth = req->rq_task->tk_msg.rpc_cred->cr_auth; uint32_t attrs[2] = { FATTR4_WORD0_RDATTR_ERROR|FATTR4_WORD0_FILEID, FATTR4_WORD1_MOUNTED_ON_FILEID, @@ -1117,7 +1154,7 @@ static int encode_readdir(struct xdr_stream *xdr, const struct nfs4_readdir_arg static int encode_readlink(struct xdr_stream *xdr, const struct nfs4_readlink *readlink, struct rpc_rqst *req) { - struct rpc_auth *auth = req->rq_task->tk_auth; + struct rpc_auth *auth = req->rq_task->tk_msg.rpc_cred->cr_auth; unsigned int replen; __be32 *p; @@ -1735,7 +1772,7 @@ out: */ static int nfs4_xdr_enc_read(struct rpc_rqst *req, __be32 *p, struct nfs_readargs *args) { - struct rpc_auth *auth = req->rq_task->tk_auth; + struct rpc_auth *auth = req->rq_task->tk_msg.rpc_cred->cr_auth; struct xdr_stream xdr; struct compound_hdr hdr = { .nops = 2, @@ -1795,7 +1832,7 @@ nfs4_xdr_enc_getacl(struct rpc_rqst *req, __be32 *p, struct nfs_getaclargs *args) { struct xdr_stream xdr; - struct rpc_auth *auth = req->rq_task->tk_auth; + struct rpc_auth *auth = req->rq_task->tk_msg.rpc_cred->cr_auth; struct compound_hdr hdr = { .nops = 2, }; @@ -2030,7 +2067,7 @@ static int nfs4_xdr_enc_fs_locations(struct rpc_rqst *req, __be32 *p, struct nfs struct compound_hdr hdr = { .nops = 3, }; - struct rpc_auth *auth = req->rq_task->tk_auth; + struct rpc_auth *auth = req->rq_task->tk_msg.rpc_cred->cr_auth; int replen; int status; @@ -3269,7 +3306,7 @@ static int decode_delegation(struct xdr_stream *xdr, struct nfs_openres *res) static int decode_open(struct xdr_stream *xdr, struct nfs_openres *res) { __be32 *p; - uint32_t bmlen; + uint32_t savewords, bmlen, i; int status; status = decode_op_hdr(xdr, OP_OPEN); @@ -3287,7 +3324,12 @@ static int decode_open(struct xdr_stream *xdr, struct nfs_openres *res) goto xdr_error; READ_BUF(bmlen << 2); - p += bmlen; + savewords = min_t(uint32_t, bmlen, NFS4_BITMAP_SIZE); + for (i = 0; i < savewords; ++i) + READ32(res->attrset[i]); + for (; i < NFS4_BITMAP_SIZE; i++) + res->attrset[i] = 0; + return decode_delegation(xdr, res); xdr_error: dprintk("%s: Bitmap too large! Length = %u\n", __FUNCTION__, bmlen); diff --git a/fs/nfs/nfsroot.c b/fs/nfs/nfsroot.c index 49d1008ce1d7..3490322d1145 100644 --- a/fs/nfs/nfsroot.c +++ b/fs/nfs/nfsroot.c @@ -428,7 +428,7 @@ static int __init root_nfs_getport(int program, int version, int proto) printk(KERN_NOTICE "Looking up port of RPC %d/%d on %u.%u.%u.%u\n", program, version, NIPQUAD(servaddr)); set_sockaddr(&sin, servaddr, 0); - return rpcb_getport_external(&sin, program, version, proto); + return rpcb_getport_sync(&sin, program, version, proto); } @@ -496,7 +496,8 @@ static int __init root_nfs_get_handle(void) NFS_MNT3_VERSION : NFS_MNT_VERSION; set_sockaddr(&sin, servaddr, htons(mount_port)); - status = nfsroot_mount(&sin, nfs_path, &fh, version, protocol); + status = nfs_mount((struct sockaddr *) &sin, sizeof(sin), NULL, + nfs_path, version, protocol, &fh); if (status < 0) printk(KERN_ERR "Root-NFS: Server returned error %d " "while mounting %s\n", status, nfs_path); diff --git a/fs/nfs/pagelist.c b/fs/nfs/pagelist.c index c5bb51a29e80..f56dae5216f4 100644 --- a/fs/nfs/pagelist.c +++ b/fs/nfs/pagelist.c @@ -85,9 +85,8 @@ nfs_create_request(struct nfs_open_context *ctx, struct inode *inode, req->wb_offset = offset; req->wb_pgbase = offset; req->wb_bytes = count; - atomic_set(&req->wb_count, 1); req->wb_context = get_nfs_open_context(ctx); - + kref_init(&req->wb_kref); return req; } @@ -109,30 +108,31 @@ void nfs_unlock_request(struct nfs_page *req) } /** - * nfs_set_page_writeback_locked - Lock a request for writeback + * nfs_set_page_tag_locked - Tag a request as locked * @req: */ -int nfs_set_page_writeback_locked(struct nfs_page *req) +static int nfs_set_page_tag_locked(struct nfs_page *req) { - struct nfs_inode *nfsi = NFS_I(req->wb_context->dentry->d_inode); + struct nfs_inode *nfsi = NFS_I(req->wb_context->path.dentry->d_inode); if (!nfs_lock_request(req)) return 0; - radix_tree_tag_set(&nfsi->nfs_page_tree, req->wb_index, NFS_PAGE_TAG_WRITEBACK); + radix_tree_tag_set(&nfsi->nfs_page_tree, req->wb_index, NFS_PAGE_TAG_LOCKED); return 1; } /** - * nfs_clear_page_writeback - Unlock request and wake up sleepers + * nfs_clear_page_tag_locked - Clear request tag and wake up sleepers */ -void nfs_clear_page_writeback(struct nfs_page *req) +void nfs_clear_page_tag_locked(struct nfs_page *req) { - struct nfs_inode *nfsi = NFS_I(req->wb_context->dentry->d_inode); + struct inode *inode = req->wb_context->path.dentry->d_inode; + struct nfs_inode *nfsi = NFS_I(inode); if (req->wb_page != NULL) { - spin_lock(&nfsi->req_lock); - radix_tree_tag_clear(&nfsi->nfs_page_tree, req->wb_index, NFS_PAGE_TAG_WRITEBACK); - spin_unlock(&nfsi->req_lock); + spin_lock(&inode->i_lock); + radix_tree_tag_clear(&nfsi->nfs_page_tree, req->wb_index, NFS_PAGE_TAG_LOCKED); + spin_unlock(&inode->i_lock); } nfs_unlock_request(req); } @@ -160,11 +160,9 @@ void nfs_clear_request(struct nfs_page *req) * * Note: Should never be called with the spinlock held! */ -void -nfs_release_request(struct nfs_page *req) +static void nfs_free_request(struct kref *kref) { - if (!atomic_dec_and_test(&req->wb_count)) - return; + struct nfs_page *req = container_of(kref, struct nfs_page, wb_kref); /* Release struct file or cached credential */ nfs_clear_request(req); @@ -172,6 +170,11 @@ nfs_release_request(struct nfs_page *req) nfs_page_free(req); } +void nfs_release_request(struct nfs_page *req) +{ + kref_put(&req->wb_kref, nfs_free_request); +} + static int nfs_wait_bit_interruptible(void *word) { int ret = 0; @@ -193,7 +196,7 @@ static int nfs_wait_bit_interruptible(void *word) int nfs_wait_on_request(struct nfs_page *req) { - struct rpc_clnt *clnt = NFS_CLIENT(req->wb_context->dentry->d_inode); + struct rpc_clnt *clnt = NFS_CLIENT(req->wb_context->path.dentry->d_inode); sigset_t oldmask; int ret = 0; @@ -379,20 +382,20 @@ void nfs_pageio_cond_complete(struct nfs_pageio_descriptor *desc, pgoff_t index) /** * nfs_scan_list - Scan a list for matching requests * @nfsi: NFS inode - * @head: One of the NFS inode request lists * @dst: Destination list * @idx_start: lower bound of page->index to scan * @npages: idx_start + npages sets the upper bound to scan. + * @tag: tag to scan for * * Moves elements from one of the inode request lists. * If the number of requests is set to 0, the entire address_space * starting at index idx_start, is scanned. * The requests are *not* checked to ensure that they form a contiguous set. - * You must be holding the inode's req_lock when calling this function + * You must be holding the inode's i_lock when calling this function */ -int nfs_scan_list(struct nfs_inode *nfsi, struct list_head *head, +int nfs_scan_list(struct nfs_inode *nfsi, struct list_head *dst, pgoff_t idx_start, - unsigned int npages) + unsigned int npages, int tag) { struct nfs_page *pgvec[NFS_SCAN_MAXENTRIES]; struct nfs_page *req; @@ -407,9 +410,9 @@ int nfs_scan_list(struct nfs_inode *nfsi, struct list_head *head, idx_end = idx_start + npages - 1; for (;;) { - found = radix_tree_gang_lookup(&nfsi->nfs_page_tree, + found = radix_tree_gang_lookup_tag(&nfsi->nfs_page_tree, (void **)&pgvec[0], idx_start, - NFS_SCAN_MAXENTRIES); + NFS_SCAN_MAXENTRIES, tag); if (found <= 0) break; for (i = 0; i < found; i++) { @@ -417,15 +420,18 @@ int nfs_scan_list(struct nfs_inode *nfsi, struct list_head *head, if (req->wb_index > idx_end) goto out; idx_start = req->wb_index + 1; - if (req->wb_list_head != head) - continue; - if (nfs_set_page_writeback_locked(req)) { + if (nfs_set_page_tag_locked(req)) { nfs_list_remove_request(req); + radix_tree_tag_clear(&nfsi->nfs_page_tree, + req->wb_index, tag); nfs_list_add_request(req, dst); res++; + if (res == INT_MAX) + goto out; } } - + /* for latency reduction */ + cond_resched_lock(&nfsi->vfs_inode.i_lock); } out: return res; diff --git a/fs/nfs/read.c b/fs/nfs/read.c index 7bd7cb95c034..6ae2e58ed05a 100644 --- a/fs/nfs/read.c +++ b/fs/nfs/read.c @@ -145,8 +145,8 @@ static void nfs_readpage_release(struct nfs_page *req) unlock_page(req->wb_page); dprintk("NFS: read done (%s/%Ld %d@%Ld)\n", - req->wb_context->dentry->d_inode->i_sb->s_id, - (long long)NFS_FILEID(req->wb_context->dentry->d_inode), + req->wb_context->path.dentry->d_inode->i_sb->s_id, + (long long)NFS_FILEID(req->wb_context->path.dentry->d_inode), req->wb_bytes, (long long)req_offset(req)); nfs_clear_request(req); @@ -164,7 +164,7 @@ static void nfs_read_rpcsetup(struct nfs_page *req, struct nfs_read_data *data, int flags; data->req = req; - data->inode = inode = req->wb_context->dentry->d_inode; + data->inode = inode = req->wb_context->path.dentry->d_inode; data->cred = req->wb_context->cred; data->args.fh = NFS_FH(inode); @@ -483,17 +483,19 @@ int nfs_readpage(struct file *file, struct page *page) */ error = nfs_wb_page(inode, page); if (error) - goto out_error; + goto out_unlock; + if (PageUptodate(page)) + goto out_unlock; error = -ESTALE; if (NFS_STALE(inode)) - goto out_error; + goto out_unlock; if (file == NULL) { error = -EBADF; ctx = nfs_find_open_context(inode, NULL, FMODE_READ); if (ctx == NULL) - goto out_error; + goto out_unlock; } else ctx = get_nfs_open_context((struct nfs_open_context *) file->private_data); @@ -502,8 +504,7 @@ int nfs_readpage(struct file *file, struct page *page) put_nfs_open_context(ctx); return error; - -out_error: +out_unlock: unlock_page(page); return error; } @@ -520,21 +521,32 @@ readpage_async_filler(void *data, struct page *page) struct inode *inode = page->mapping->host; struct nfs_page *new; unsigned int len; + int error; + + error = nfs_wb_page(inode, page); + if (error) + goto out_unlock; + if (PageUptodate(page)) + goto out_unlock; - nfs_wb_page(inode, page); len = nfs_page_length(page); if (len == 0) return nfs_return_empty_page(page); + new = nfs_create_request(desc->ctx, inode, page, 0, len); - if (IS_ERR(new)) { - SetPageError(page); - unlock_page(page); - return PTR_ERR(new); - } + if (IS_ERR(new)) + goto out_error; + if (len < PAGE_CACHE_SIZE) zero_user_page(page, len, PAGE_CACHE_SIZE - len, KM_USER0); nfs_pageio_add_request(desc->pgio, new); return 0; +out_error: + error = PTR_ERR(new); + SetPageError(page); +out_unlock: + unlock_page(page); + return error; } int nfs_readpages(struct file *filp, struct address_space *mapping, diff --git a/fs/nfs/super.c b/fs/nfs/super.c index ca20d3cc2609..a2b1af89ca1a 100644 --- a/fs/nfs/super.c +++ b/fs/nfs/super.c @@ -45,6 +45,7 @@ #include <linux/inet.h> #include <linux/nfs_xdr.h> #include <linux/magic.h> +#include <linux/parser.h> #include <asm/system.h> #include <asm/uaccess.h> @@ -57,6 +58,167 @@ #define NFSDBG_FACILITY NFSDBG_VFS + +struct nfs_parsed_mount_data { + int flags; + int rsize, wsize; + int timeo, retrans; + int acregmin, acregmax, + acdirmin, acdirmax; + int namlen; + unsigned int bsize; + unsigned int auth_flavor_len; + rpc_authflavor_t auth_flavors[1]; + char *client_address; + + struct { + struct sockaddr_in address; + unsigned int program; + unsigned int version; + unsigned short port; + int protocol; + } mount_server; + + struct { + struct sockaddr_in address; + char *hostname; + char *export_path; + unsigned int program; + int protocol; + } nfs_server; +}; + +enum { + /* Mount options that take no arguments */ + Opt_soft, Opt_hard, + Opt_intr, Opt_nointr, + Opt_posix, Opt_noposix, + Opt_cto, Opt_nocto, + Opt_ac, Opt_noac, + Opt_lock, Opt_nolock, + Opt_v2, Opt_v3, + Opt_udp, Opt_tcp, + Opt_acl, Opt_noacl, + Opt_rdirplus, Opt_nordirplus, + Opt_sharecache, Opt_nosharecache, + + /* Mount options that take integer arguments */ + Opt_port, + Opt_rsize, Opt_wsize, Opt_bsize, + Opt_timeo, Opt_retrans, + Opt_acregmin, Opt_acregmax, + Opt_acdirmin, Opt_acdirmax, + Opt_actimeo, + Opt_namelen, + Opt_mountport, + Opt_mountprog, Opt_mountvers, + Opt_nfsprog, Opt_nfsvers, + + /* Mount options that take string arguments */ + Opt_sec, Opt_proto, Opt_mountproto, + Opt_addr, Opt_mounthost, Opt_clientaddr, + + /* Mount options that are ignored */ + Opt_userspace, Opt_deprecated, + + Opt_err +}; + +static match_table_t nfs_mount_option_tokens = { + { Opt_userspace, "bg" }, + { Opt_userspace, "fg" }, + { Opt_soft, "soft" }, + { Opt_hard, "hard" }, + { Opt_intr, "intr" }, + { Opt_nointr, "nointr" }, + { Opt_posix, "posix" }, + { Opt_noposix, "noposix" }, + { Opt_cto, "cto" }, + { Opt_nocto, "nocto" }, + { Opt_ac, "ac" }, + { Opt_noac, "noac" }, + { Opt_lock, "lock" }, + { Opt_nolock, "nolock" }, + { Opt_v2, "v2" }, + { Opt_v3, "v3" }, + { Opt_udp, "udp" }, + { Opt_tcp, "tcp" }, + { Opt_acl, "acl" }, + { Opt_noacl, "noacl" }, + { Opt_rdirplus, "rdirplus" }, + { Opt_nordirplus, "nordirplus" }, + { Opt_sharecache, "sharecache" }, + { Opt_nosharecache, "nosharecache" }, + + { Opt_port, "port=%u" }, + { Opt_rsize, "rsize=%u" }, + { Opt_wsize, "wsize=%u" }, + { Opt_bsize, "bsize=%u" }, + { Opt_timeo, "timeo=%u" }, + { Opt_retrans, "retrans=%u" }, + { Opt_acregmin, "acregmin=%u" }, + { Opt_acregmax, "acregmax=%u" }, + { Opt_acdirmin, "acdirmin=%u" }, + { Opt_acdirmax, "acdirmax=%u" }, + { Opt_actimeo, "actimeo=%u" }, + { Opt_userspace, "retry=%u" }, + { Opt_namelen, "namlen=%u" }, + { Opt_mountport, "mountport=%u" }, + { Opt_mountprog, "mountprog=%u" }, + { Opt_mountvers, "mountvers=%u" }, + { Opt_nfsprog, "nfsprog=%u" }, + { Opt_nfsvers, "nfsvers=%u" }, + { Opt_nfsvers, "vers=%u" }, + + { Opt_sec, "sec=%s" }, + { Opt_proto, "proto=%s" }, + { Opt_mountproto, "mountproto=%s" }, + { Opt_addr, "addr=%s" }, + { Opt_clientaddr, "clientaddr=%s" }, + { Opt_mounthost, "mounthost=%s" }, + + { Opt_err, NULL } +}; + +enum { + Opt_xprt_udp, Opt_xprt_tcp, + + Opt_xprt_err +}; + +static match_table_t nfs_xprt_protocol_tokens = { + { Opt_xprt_udp, "udp" }, + { Opt_xprt_tcp, "tcp" }, + + { Opt_xprt_err, NULL } +}; + +enum { + Opt_sec_none, Opt_sec_sys, + Opt_sec_krb5, Opt_sec_krb5i, Opt_sec_krb5p, + Opt_sec_lkey, Opt_sec_lkeyi, Opt_sec_lkeyp, + Opt_sec_spkm, Opt_sec_spkmi, Opt_sec_spkmp, + + Opt_sec_err +}; + +static match_table_t nfs_secflavor_tokens = { + { Opt_sec_none, "none" }, + { Opt_sec_none, "null" }, + { Opt_sec_sys, "sys" }, + + { Opt_sec_krb5, "krb5" }, + { Opt_sec_krb5i, "krb5i" }, + { Opt_sec_krb5p, "krb5p" }, + + { Opt_sec_lkey, "lkey" }, + { Opt_sec_lkeyi, "lkeyi" }, + { Opt_sec_lkeyp, "lkeyp" }, + + { Opt_sec_err, NULL } +}; + + static void nfs_umount_begin(struct vfsmount *, int); static int nfs_statfs(struct dentry *, struct kstatfs *); static int nfs_show_options(struct seq_file *, struct vfsmount *); @@ -263,11 +425,11 @@ static const char *nfs_pseudoflavour_to_name(rpc_authflavor_t flavour) { RPC_AUTH_GSS_SPKM, "spkm" }, { RPC_AUTH_GSS_SPKMI, "spkmi" }, { RPC_AUTH_GSS_SPKMP, "spkmp" }, - { -1, "unknown" } + { UINT_MAX, "unknown" } }; int i; - for (i=0; sec_flavours[i].flavour != -1; i++) { + for (i = 0; sec_flavours[i].flavour != UINT_MAX; i++) { if (sec_flavours[i].flavour == flavour) break; } @@ -291,6 +453,7 @@ static void nfs_show_mount_options(struct seq_file *m, struct nfs_server *nfss, { NFS_MOUNT_NONLM, ",nolock", "" }, { NFS_MOUNT_NOACL, ",noacl", "" }, { NFS_MOUNT_NORDIRPLUS, ",nordirplus", "" }, + { NFS_MOUNT_UNSHARED, ",nosharecache", ""}, { 0, NULL, NULL } }; const struct proc_nfs_info *nfs_infop; @@ -430,87 +593,641 @@ static int nfs_show_stats(struct seq_file *m, struct vfsmount *mnt) */ static void nfs_umount_begin(struct vfsmount *vfsmnt, int flags) { + struct nfs_server *server = NFS_SB(vfsmnt->mnt_sb); + struct rpc_clnt *rpc; + shrink_submounts(vfsmnt, &nfs_automount_list); + + if (!(flags & MNT_FORCE)) + return; + /* -EIO all pending I/O */ + rpc = server->client_acl; + if (!IS_ERR(rpc)) + rpc_killall_tasks(rpc); + rpc = server->client; + if (!IS_ERR(rpc)) + rpc_killall_tasks(rpc); } /* - * Validate the NFS2/NFS3 mount data - * - fills in the mount root filehandle + * Sanity-check a server address provided by the mount command */ -static int nfs_validate_mount_data(struct nfs_mount_data *data, - struct nfs_fh *mntfh) +static int nfs_verify_server_address(struct sockaddr *addr) { - if (data == NULL) { - dprintk("%s: missing data argument\n", __FUNCTION__); - return -EINVAL; + switch (addr->sa_family) { + case AF_INET: { + struct sockaddr_in *sa = (struct sockaddr_in *) addr; + if (sa->sin_addr.s_addr != INADDR_ANY) + return 1; + break; + } } - if (data->version <= 0 || data->version > NFS_MOUNT_VERSION) { - dprintk("%s: bad mount version\n", __FUNCTION__); - return -EINVAL; + return 0; +} + +/* + * Error-check and convert a string of mount options from user space into + * a data structure + */ +static int nfs_parse_mount_options(char *raw, + struct nfs_parsed_mount_data *mnt) +{ + char *p, *string; + + if (!raw) { + dfprintk(MOUNT, "NFS: mount options string was NULL.\n"); + return 1; } + dfprintk(MOUNT, "NFS: nfs mount opts='%s'\n", raw); - switch (data->version) { - case 1: - data->namlen = 0; - case 2: - data->bsize = 0; - case 3: - if (data->flags & NFS_MOUNT_VER3) { - dprintk("%s: mount structure version %d does not support NFSv3\n", - __FUNCTION__, - data->version); - return -EINVAL; + while ((p = strsep(&raw, ",")) != NULL) { + substring_t args[MAX_OPT_ARGS]; + int option, token; + + if (!*p) + continue; + + dfprintk(MOUNT, "NFS: parsing nfs mount option '%s'\n", p); + + token = match_token(p, nfs_mount_option_tokens, args); + switch (token) { + case Opt_soft: + mnt->flags |= NFS_MOUNT_SOFT; + break; + case Opt_hard: + mnt->flags &= ~NFS_MOUNT_SOFT; + break; + case Opt_intr: + mnt->flags |= NFS_MOUNT_INTR; + break; + case Opt_nointr: + mnt->flags &= ~NFS_MOUNT_INTR; + break; + case Opt_posix: + mnt->flags |= NFS_MOUNT_POSIX; + break; + case Opt_noposix: + mnt->flags &= ~NFS_MOUNT_POSIX; + break; + case Opt_cto: + mnt->flags &= ~NFS_MOUNT_NOCTO; + break; + case Opt_nocto: + mnt->flags |= NFS_MOUNT_NOCTO; + break; + case Opt_ac: + mnt->flags &= ~NFS_MOUNT_NOAC; + break; + case Opt_noac: + mnt->flags |= NFS_MOUNT_NOAC; + break; + case Opt_lock: + mnt->flags &= ~NFS_MOUNT_NONLM; + break; + case Opt_nolock: + mnt->flags |= NFS_MOUNT_NONLM; + break; + case Opt_v2: + mnt->flags &= ~NFS_MOUNT_VER3; + break; + case Opt_v3: + mnt->flags |= NFS_MOUNT_VER3; + break; + case Opt_udp: + mnt->flags &= ~NFS_MOUNT_TCP; + mnt->nfs_server.protocol = IPPROTO_UDP; + mnt->timeo = 7; + mnt->retrans = 5; + break; + case Opt_tcp: + mnt->flags |= NFS_MOUNT_TCP; + mnt->nfs_server.protocol = IPPROTO_TCP; + mnt->timeo = 600; + mnt->retrans = 2; + break; + case Opt_acl: + mnt->flags &= ~NFS_MOUNT_NOACL; + break; + case Opt_noacl: + mnt->flags |= NFS_MOUNT_NOACL; + break; + case Opt_rdirplus: + mnt->flags &= ~NFS_MOUNT_NORDIRPLUS; + break; + case Opt_nordirplus: + mnt->flags |= NFS_MOUNT_NORDIRPLUS; + break; + case Opt_sharecache: + mnt->flags &= ~NFS_MOUNT_UNSHARED; + break; + case Opt_nosharecache: + mnt->flags |= NFS_MOUNT_UNSHARED; + break; + + case Opt_port: + if (match_int(args, &option)) + return 0; + if (option < 0 || option > 65535) + return 0; + mnt->nfs_server.address.sin_port = htonl(option); + break; + case Opt_rsize: + if (match_int(args, &mnt->rsize)) + return 0; + break; + case Opt_wsize: + if (match_int(args, &mnt->wsize)) + return 0; + break; + case Opt_bsize: + if (match_int(args, &option)) + return 0; + if (option < 0) + return 0; + mnt->bsize = option; + break; + case Opt_timeo: + if (match_int(args, &mnt->timeo)) + return 0; + break; + case Opt_retrans: + if (match_int(args, &mnt->retrans)) + return 0; + break; + case Opt_acregmin: + if (match_int(args, &mnt->acregmin)) + return 0; + break; + case Opt_acregmax: + if (match_int(args, &mnt->acregmax)) + return 0; + break; + case Opt_acdirmin: + if (match_int(args, &mnt->acdirmin)) + return 0; + break; + case Opt_acdirmax: + if (match_int(args, &mnt->acdirmax)) + return 0; + break; + case Opt_actimeo: + if (match_int(args, &option)) + return 0; + if (option < 0) + return 0; + mnt->acregmin = + mnt->acregmax = + mnt->acdirmin = + mnt->acdirmax = option; + break; + case Opt_namelen: + if (match_int(args, &mnt->namlen)) + return 0; + break; + case Opt_mountport: + if (match_int(args, &option)) + return 0; + if (option < 0 || option > 65535) + return 0; + mnt->mount_server.port = option; + break; + case Opt_mountprog: + if (match_int(args, &option)) + return 0; + if (option < 0) + return 0; + mnt->mount_server.program = option; + break; + case Opt_mountvers: + if (match_int(args, &option)) + return 0; + if (option < 0) + return 0; + mnt->mount_server.version = option; + break; + case Opt_nfsprog: + if (match_int(args, &option)) + return 0; + if (option < 0) + return 0; + mnt->nfs_server.program = option; + break; + case Opt_nfsvers: + if (match_int(args, &option)) + return 0; + switch (option) { + case 2: + mnt->flags &= ~NFS_MOUNT_VER3; + break; + case 3: + mnt->flags |= NFS_MOUNT_VER3; + break; + default: + goto out_unrec_vers; } - data->root.size = NFS2_FHSIZE; - memcpy(data->root.data, data->old_root.data, NFS2_FHSIZE); - case 4: - if (data->flags & NFS_MOUNT_SECFLAVOUR) { - dprintk("%s: mount structure version %d does not support strong security\n", - __FUNCTION__, - data->version); - return -EINVAL; + break; + + case Opt_sec: + string = match_strdup(args); + if (string == NULL) + goto out_nomem; + token = match_token(string, nfs_secflavor_tokens, args); + kfree(string); + + /* + * The flags setting is for v2/v3. The flavor_len + * setting is for v4. v2/v3 also need to know the + * difference between NULL and UNIX. + */ + switch (token) { + case Opt_sec_none: + mnt->flags &= ~NFS_MOUNT_SECFLAVOUR; + mnt->auth_flavor_len = 0; + mnt->auth_flavors[0] = RPC_AUTH_NULL; + break; + case Opt_sec_sys: + mnt->flags &= ~NFS_MOUNT_SECFLAVOUR; + mnt->auth_flavor_len = 0; + mnt->auth_flavors[0] = RPC_AUTH_UNIX; + break; + case Opt_sec_krb5: + mnt->flags |= NFS_MOUNT_SECFLAVOUR; + mnt->auth_flavor_len = 1; + mnt->auth_flavors[0] = RPC_AUTH_GSS_KRB5; + break; + case Opt_sec_krb5i: + mnt->flags |= NFS_MOUNT_SECFLAVOUR; + mnt->auth_flavor_len = 1; + mnt->auth_flavors[0] = RPC_AUTH_GSS_KRB5I; + break; + case Opt_sec_krb5p: + mnt->flags |= NFS_MOUNT_SECFLAVOUR; + mnt->auth_flavor_len = 1; + mnt->auth_flavors[0] = RPC_AUTH_GSS_KRB5P; + break; + case Opt_sec_lkey: + mnt->flags |= NFS_MOUNT_SECFLAVOUR; + mnt->auth_flavor_len = 1; + mnt->auth_flavors[0] = RPC_AUTH_GSS_LKEY; + break; + case Opt_sec_lkeyi: + mnt->flags |= NFS_MOUNT_SECFLAVOUR; + mnt->auth_flavor_len = 1; + mnt->auth_flavors[0] = RPC_AUTH_GSS_LKEYI; + break; + case Opt_sec_lkeyp: + mnt->flags |= NFS_MOUNT_SECFLAVOUR; + mnt->auth_flavor_len = 1; + mnt->auth_flavors[0] = RPC_AUTH_GSS_LKEYP; + break; + case Opt_sec_spkm: + mnt->flags |= NFS_MOUNT_SECFLAVOUR; + mnt->auth_flavor_len = 1; + mnt->auth_flavors[0] = RPC_AUTH_GSS_SPKM; + break; + case Opt_sec_spkmi: + mnt->flags |= NFS_MOUNT_SECFLAVOUR; + mnt->auth_flavor_len = 1; + mnt->auth_flavors[0] = RPC_AUTH_GSS_SPKMI; + break; + case Opt_sec_spkmp: + mnt->flags |= NFS_MOUNT_SECFLAVOUR; + mnt->auth_flavor_len = 1; + mnt->auth_flavors[0] = RPC_AUTH_GSS_SPKMP; + break; + default: + goto out_unrec_sec; } - case 5: - memset(data->context, 0, sizeof(data->context)); - } + break; + case Opt_proto: + string = match_strdup(args); + if (string == NULL) + goto out_nomem; + token = match_token(string, + nfs_xprt_protocol_tokens, args); + kfree(string); + + switch (token) { + case Opt_udp: + mnt->flags &= ~NFS_MOUNT_TCP; + mnt->nfs_server.protocol = IPPROTO_UDP; + mnt->timeo = 7; + mnt->retrans = 5; + break; + case Opt_tcp: + mnt->flags |= NFS_MOUNT_TCP; + mnt->nfs_server.protocol = IPPROTO_TCP; + mnt->timeo = 600; + mnt->retrans = 2; + break; + default: + goto out_unrec_xprt; + } + break; + case Opt_mountproto: + string = match_strdup(args); + if (string == NULL) + goto out_nomem; + token = match_token(string, + nfs_xprt_protocol_tokens, args); + kfree(string); + + switch (token) { + case Opt_udp: + mnt->mount_server.protocol = IPPROTO_UDP; + break; + case Opt_tcp: + mnt->mount_server.protocol = IPPROTO_TCP; + break; + default: + goto out_unrec_xprt; + } + break; + case Opt_addr: + string = match_strdup(args); + if (string == NULL) + goto out_nomem; + mnt->nfs_server.address.sin_family = AF_INET; + mnt->nfs_server.address.sin_addr.s_addr = + in_aton(string); + kfree(string); + break; + case Opt_clientaddr: + string = match_strdup(args); + if (string == NULL) + goto out_nomem; + mnt->client_address = string; + break; + case Opt_mounthost: + string = match_strdup(args); + if (string == NULL) + goto out_nomem; + mnt->mount_server.address.sin_family = AF_INET; + mnt->mount_server.address.sin_addr.s_addr = + in_aton(string); + kfree(string); + break; - /* Set the pseudoflavor */ - if (!(data->flags & NFS_MOUNT_SECFLAVOUR)) - data->pseudoflavor = RPC_AUTH_UNIX; + case Opt_userspace: + case Opt_deprecated: + break; -#ifndef CONFIG_NFS_V3 - /* If NFSv3 is not compiled in, return -EPROTONOSUPPORT */ - if (data->flags & NFS_MOUNT_VER3) { - dprintk("%s: NFSv3 not compiled into kernel\n", __FUNCTION__); - return -EPROTONOSUPPORT; + default: + goto out_unknown; + } } -#endif /* CONFIG_NFS_V3 */ - /* We now require that the mount process passes the remote address */ - if (data->addr.sin_addr.s_addr == INADDR_ANY) { - dprintk("%s: mount program didn't pass remote address!\n", - __FUNCTION__); - return -EINVAL; + return 1; + +out_nomem: + printk(KERN_INFO "NFS: not enough memory to parse option\n"); + return 0; + +out_unrec_vers: + printk(KERN_INFO "NFS: unrecognized NFS version number\n"); + return 0; + +out_unrec_xprt: + printk(KERN_INFO "NFS: unrecognized transport protocol\n"); + return 0; + +out_unrec_sec: + printk(KERN_INFO "NFS: unrecognized security flavor\n"); + return 0; + +out_unknown: + printk(KERN_INFO "NFS: unknown mount option: %s\n", p); + return 0; +} + +/* + * Use the remote server's MOUNT service to request the NFS file handle + * corresponding to the provided path. + */ +static int nfs_try_mount(struct nfs_parsed_mount_data *args, + struct nfs_fh *root_fh) +{ + struct sockaddr_in sin; + int status; + + if (args->mount_server.version == 0) { + if (args->flags & NFS_MOUNT_VER3) + args->mount_server.version = NFS_MNT3_VERSION; + else + args->mount_server.version = NFS_MNT_VERSION; } - /* Prepare the root filehandle */ - if (data->flags & NFS_MOUNT_VER3) - mntfh->size = data->root.size; + /* + * Construct the mount server's address. + */ + if (args->mount_server.address.sin_addr.s_addr != INADDR_ANY) + sin = args->mount_server.address; else - mntfh->size = NFS2_FHSIZE; + sin = args->nfs_server.address; + if (args->mount_server.port == 0) { + status = rpcb_getport_sync(&sin, + args->mount_server.program, + args->mount_server.version, + args->mount_server.protocol); + if (status < 0) + goto out_err; + sin.sin_port = htons(status); + } else + sin.sin_port = htons(args->mount_server.port); + + /* + * Now ask the mount server to map our export path + * to a file handle. + */ + status = nfs_mount((struct sockaddr *) &sin, + sizeof(sin), + args->nfs_server.hostname, + args->nfs_server.export_path, + args->mount_server.version, + args->mount_server.protocol, + root_fh); + if (status < 0) + goto out_err; + + return status; - if (mntfh->size > sizeof(mntfh->data)) { - dprintk("%s: invalid root filehandle\n", __FUNCTION__); - return -EINVAL; +out_err: + dfprintk(MOUNT, "NFS: unable to contact server on host " + NIPQUAD_FMT "\n", NIPQUAD(sin.sin_addr.s_addr)); + return status; +} + +/* + * Validate the NFS2/NFS3 mount data + * - fills in the mount root filehandle + * + * For option strings, user space handles the following behaviors: + * + * + DNS: mapping server host name to IP address ("addr=" option) + * + * + failure mode: how to behave if a mount request can't be handled + * immediately ("fg/bg" option) + * + * + retry: how often to retry a mount request ("retry=" option) + * + * + breaking back: trying proto=udp after proto=tcp, v2 after v3, + * mountproto=tcp after mountproto=udp, and so on + * + * XXX: as far as I can tell, changing the NFS program number is not + * supported in the NFS client. + */ +static int nfs_validate_mount_data(struct nfs_mount_data **options, + struct nfs_fh *mntfh, + const char *dev_name) +{ + struct nfs_mount_data *data = *options; + + if (data == NULL) + goto out_no_data; + + switch (data->version) { + case 1: + data->namlen = 0; + case 2: + data->bsize = 0; + case 3: + if (data->flags & NFS_MOUNT_VER3) + goto out_no_v3; + data->root.size = NFS2_FHSIZE; + memcpy(data->root.data, data->old_root.data, NFS2_FHSIZE); + case 4: + if (data->flags & NFS_MOUNT_SECFLAVOUR) + goto out_no_sec; + case 5: + memset(data->context, 0, sizeof(data->context)); + case 6: + if (data->flags & NFS_MOUNT_VER3) + mntfh->size = data->root.size; + else + mntfh->size = NFS2_FHSIZE; + + if (mntfh->size > sizeof(mntfh->data)) + goto out_invalid_fh; + + memcpy(mntfh->data, data->root.data, mntfh->size); + if (mntfh->size < sizeof(mntfh->data)) + memset(mntfh->data + mntfh->size, 0, + sizeof(mntfh->data) - mntfh->size); + break; + default: { + unsigned int len; + char *c; + int status; + struct nfs_parsed_mount_data args = { + .flags = (NFS_MOUNT_VER3 | NFS_MOUNT_TCP), + .rsize = NFS_MAX_FILE_IO_SIZE, + .wsize = NFS_MAX_FILE_IO_SIZE, + .timeo = 600, + .retrans = 2, + .acregmin = 3, + .acregmax = 60, + .acdirmin = 30, + .acdirmax = 60, + .mount_server.protocol = IPPROTO_UDP, + .mount_server.program = NFS_MNT_PROGRAM, + .nfs_server.protocol = IPPROTO_TCP, + .nfs_server.program = NFS_PROGRAM, + }; + + if (nfs_parse_mount_options((char *) *options, &args) == 0) + return -EINVAL; + + data = kzalloc(sizeof(*data), GFP_KERNEL); + if (data == NULL) + return -ENOMEM; + + /* + * NB: after this point, caller will free "data" + * if we return an error + */ + *options = data; + + c = strchr(dev_name, ':'); + if (c == NULL) + return -EINVAL; + len = c - dev_name - 1; + if (len > sizeof(data->hostname)) + return -EINVAL; + strncpy(data->hostname, dev_name, len); + args.nfs_server.hostname = data->hostname; + + c++; + if (strlen(c) > NFS_MAXPATHLEN) + return -EINVAL; + args.nfs_server.export_path = c; + + status = nfs_try_mount(&args, mntfh); + if (status) + return -EINVAL; + + /* + * Translate to nfs_mount_data, which nfs_fill_super + * can deal with. + */ + data->version = 6; + data->flags = args.flags; + data->rsize = args.rsize; + data->wsize = args.wsize; + data->timeo = args.timeo; + data->retrans = args.retrans; + data->acregmin = args.acregmin; + data->acregmax = args.acregmax; + data->acdirmin = args.acdirmin; + data->acdirmax = args.acdirmax; + data->addr = args.nfs_server.address; + data->namlen = args.namlen; + data->bsize = args.bsize; + data->pseudoflavor = args.auth_flavors[0]; + + break; + } } - memcpy(mntfh->data, data->root.data, mntfh->size); - if (mntfh->size < sizeof(mntfh->data)) - memset(mntfh->data + mntfh->size, 0, - sizeof(mntfh->data) - mntfh->size); + if (!(data->flags & NFS_MOUNT_SECFLAVOUR)) + data->pseudoflavor = RPC_AUTH_UNIX; + +#ifndef CONFIG_NFS_V3 + if (data->flags & NFS_MOUNT_VER3) + goto out_v3_not_compiled; +#endif /* !CONFIG_NFS_V3 */ + + if (!nfs_verify_server_address((struct sockaddr *) &data->addr)) + goto out_no_address; return 0; + +out_no_data: + dfprintk(MOUNT, "NFS: mount program didn't pass any mount data\n"); + return -EINVAL; + +out_no_v3: + dfprintk(MOUNT, "NFS: nfs_mount_data version %d does not support v3\n", + data->version); + return -EINVAL; + +out_no_sec: + dfprintk(MOUNT, "NFS: nfs_mount_data version supports only AUTH_SYS\n"); + return -EINVAL; + +#ifndef CONFIG_NFS_V3 +out_v3_not_compiled: + dfprintk(MOUNT, "NFS: NFSv3 is not compiled into kernel\n"); + return -EPROTONOSUPPORT; +#endif /* !CONFIG_NFS_V3 */ + +out_no_address: + dfprintk(MOUNT, "NFS: mount program didn't pass remote address\n"); + return -EINVAL; + +out_invalid_fh: + dfprintk(MOUNT, "NFS: invalid root filehandle\n"); + return -EINVAL; } /* @@ -600,13 +1317,51 @@ static int nfs_compare_super(struct super_block *sb, void *data) { struct nfs_server *server = data, *old = NFS_SB(sb); - if (old->nfs_client != server->nfs_client) + if (memcmp(&old->nfs_client->cl_addr, + &server->nfs_client->cl_addr, + sizeof(old->nfs_client->cl_addr)) != 0) + return 0; + /* Note: NFS_MOUNT_UNSHARED == NFS4_MOUNT_UNSHARED */ + if (old->flags & NFS_MOUNT_UNSHARED) return 0; if (memcmp(&old->fsid, &server->fsid, sizeof(old->fsid)) != 0) return 0; return 1; } +#define NFS_MS_MASK (MS_RDONLY|MS_NOSUID|MS_NODEV|MS_NOEXEC|MS_SYNCHRONOUS) + +static int nfs_compare_mount_options(const struct super_block *s, const struct nfs_server *b, int flags) +{ + const struct nfs_server *a = s->s_fs_info; + const struct rpc_clnt *clnt_a = a->client; + const struct rpc_clnt *clnt_b = b->client; + + if ((s->s_flags & NFS_MS_MASK) != (flags & NFS_MS_MASK)) + goto Ebusy; + if (a->nfs_client != b->nfs_client) + goto Ebusy; + if (a->flags != b->flags) + goto Ebusy; + if (a->wsize != b->wsize) + goto Ebusy; + if (a->rsize != b->rsize) + goto Ebusy; + if (a->acregmin != b->acregmin) + goto Ebusy; + if (a->acregmax != b->acregmax) + goto Ebusy; + if (a->acdirmin != b->acdirmin) + goto Ebusy; + if (a->acdirmax != b->acdirmax) + goto Ebusy; + if (clnt_a->cl_auth->au_flavor != clnt_b->cl_auth->au_flavor) + goto Ebusy; + return 0; +Ebusy: + return -EBUSY; +} + static int nfs_get_sb(struct file_system_type *fs_type, int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt) { @@ -615,30 +1370,37 @@ static int nfs_get_sb(struct file_system_type *fs_type, struct nfs_fh mntfh; struct nfs_mount_data *data = raw_data; struct dentry *mntroot; + int (*compare_super)(struct super_block *, void *) = nfs_compare_super; int error; /* Validate the mount data */ - error = nfs_validate_mount_data(data, &mntfh); + error = nfs_validate_mount_data(&data, &mntfh, dev_name); if (error < 0) - return error; + goto out; /* Get a volume representation */ server = nfs_create_server(data, &mntfh); if (IS_ERR(server)) { error = PTR_ERR(server); - goto out_err_noserver; + goto out; } + if (server->flags & NFS_MOUNT_UNSHARED) + compare_super = NULL; + /* Get a superblock - note that we may end up sharing one that already exists */ - s = sget(fs_type, nfs_compare_super, nfs_set_super, server); + s = sget(fs_type, compare_super, nfs_set_super, server); if (IS_ERR(s)) { error = PTR_ERR(s); goto out_err_nosb; } if (s->s_fs_info != server) { + error = nfs_compare_mount_options(s, server, flags); nfs_free_server(server); server = NULL; + if (error < 0) + goto error_splat_super; } if (!s->s_root) { @@ -656,17 +1418,21 @@ static int nfs_get_sb(struct file_system_type *fs_type, s->s_flags |= MS_ACTIVE; mnt->mnt_sb = s; mnt->mnt_root = mntroot; - return 0; + error = 0; + +out: + if (data != raw_data) + kfree(data); + return error; out_err_nosb: nfs_free_server(server); -out_err_noserver: - return error; + goto out; error_splat_super: up_write(&s->s_umount); deactivate_super(s); - return error; + goto out; } /* @@ -691,6 +1457,7 @@ static int nfs_xdev_get_sb(struct file_system_type *fs_type, int flags, struct super_block *s; struct nfs_server *server; struct dentry *mntroot; + int (*compare_super)(struct super_block *, void *) = nfs_compare_super; int error; dprintk("--> nfs_xdev_get_sb()\n"); @@ -702,16 +1469,22 @@ static int nfs_xdev_get_sb(struct file_system_type *fs_type, int flags, goto out_err_noserver; } + if (server->flags & NFS_MOUNT_UNSHARED) + compare_super = NULL; + /* Get a superblock - note that we may end up sharing one that already exists */ - s = sget(&nfs_fs_type, nfs_compare_super, nfs_set_super, server); + s = sget(&nfs_fs_type, compare_super, nfs_set_super, server); if (IS_ERR(s)) { error = PTR_ERR(s); goto out_err_nosb; } if (s->s_fs_info != server) { + error = nfs_compare_mount_options(s, server, flags); nfs_free_server(server); server = NULL; + if (error < 0) + goto error_splat_super; } if (!s->s_root) { @@ -772,25 +1545,164 @@ static void nfs4_fill_super(struct super_block *sb) nfs_initialise_sb(sb); } -static void *nfs_copy_user_string(char *dst, struct nfs_string *src, int maxlen) +/* + * Validate NFSv4 mount options + */ +static int nfs4_validate_mount_data(struct nfs4_mount_data **options, + const char *dev_name, + struct sockaddr_in *addr, + rpc_authflavor_t *authflavour, + char **hostname, + char **mntpath, + char **ip_addr) { - void *p = NULL; - - if (!src->len) - return ERR_PTR(-EINVAL); - if (src->len < maxlen) - maxlen = src->len; - if (dst == NULL) { - p = dst = kmalloc(maxlen + 1, GFP_KERNEL); - if (p == NULL) - return ERR_PTR(-ENOMEM); - } - if (copy_from_user(dst, src->data, maxlen)) { - kfree(p); - return ERR_PTR(-EFAULT); + struct nfs4_mount_data *data = *options; + char *c; + + if (data == NULL) + goto out_no_data; + + switch (data->version) { + case 1: + if (data->host_addrlen != sizeof(*addr)) + goto out_no_address; + if (copy_from_user(addr, data->host_addr, sizeof(*addr))) + return -EFAULT; + if (addr->sin_port == 0) + addr->sin_port = htons(NFS_PORT); + if (!nfs_verify_server_address((struct sockaddr *) addr)) + goto out_no_address; + + switch (data->auth_flavourlen) { + case 0: + *authflavour = RPC_AUTH_UNIX; + break; + case 1: + if (copy_from_user(authflavour, data->auth_flavours, + sizeof(*authflavour))) + return -EFAULT; + break; + default: + goto out_inval_auth; + } + + c = strndup_user(data->hostname.data, NFS4_MAXNAMLEN); + if (IS_ERR(c)) + return PTR_ERR(c); + *hostname = c; + + c = strndup_user(data->mnt_path.data, NFS4_MAXPATHLEN); + if (IS_ERR(c)) + return PTR_ERR(c); + *mntpath = c; + dfprintk(MOUNT, "NFS: MNTPATH: '%s'\n", *mntpath); + + c = strndup_user(data->client_addr.data, 16); + if (IS_ERR(c)) + return PTR_ERR(c); + *ip_addr = c; + + break; + default: { + unsigned int len; + struct nfs_parsed_mount_data args = { + .rsize = NFS_MAX_FILE_IO_SIZE, + .wsize = NFS_MAX_FILE_IO_SIZE, + .timeo = 600, + .retrans = 2, + .acregmin = 3, + .acregmax = 60, + .acdirmin = 30, + .acdirmax = 60, + .nfs_server.protocol = IPPROTO_TCP, + }; + + if (nfs_parse_mount_options((char *) *options, &args) == 0) + return -EINVAL; + + if (!nfs_verify_server_address((struct sockaddr *) + &args.nfs_server.address)) + return -EINVAL; + *addr = args.nfs_server.address; + + switch (args.auth_flavor_len) { + case 0: + *authflavour = RPC_AUTH_UNIX; + break; + case 1: + *authflavour = (rpc_authflavor_t) args.auth_flavors[0]; + break; + default: + goto out_inval_auth; + } + + /* + * Translate to nfs4_mount_data, which nfs4_fill_super + * can deal with. + */ + data = kzalloc(sizeof(*data), GFP_KERNEL); + if (data == NULL) + return -ENOMEM; + *options = data; + + data->version = 1; + data->flags = args.flags & NFS4_MOUNT_FLAGMASK; + data->rsize = args.rsize; + data->wsize = args.wsize; + data->timeo = args.timeo; + data->retrans = args.retrans; + data->acregmin = args.acregmin; + data->acregmax = args.acregmax; + data->acdirmin = args.acdirmin; + data->acdirmax = args.acdirmax; + data->proto = args.nfs_server.protocol; + + /* + * Split "dev_name" into "hostname:mntpath". + */ + c = strchr(dev_name, ':'); + if (c == NULL) + return -EINVAL; + /* while calculating len, pretend ':' is '\0' */ + len = c - dev_name; + if (len > NFS4_MAXNAMLEN) + return -EINVAL; + *hostname = kzalloc(len, GFP_KERNEL); + if (*hostname == NULL) + return -ENOMEM; + strncpy(*hostname, dev_name, len - 1); + + c++; /* step over the ':' */ + len = strlen(c); + if (len > NFS4_MAXPATHLEN) + return -EINVAL; + *mntpath = kzalloc(len + 1, GFP_KERNEL); + if (*mntpath == NULL) + return -ENOMEM; + strncpy(*mntpath, c, len); + + dprintk("MNTPATH: %s\n", *mntpath); + + *ip_addr = args.client_address; + + break; + } } - dst[maxlen] = '\0'; - return dst; + + return 0; + +out_no_data: + dfprintk(MOUNT, "NFS4: mount program didn't pass any mount data\n"); + return -EINVAL; + +out_inval_auth: + dfprintk(MOUNT, "NFS4: Invalid number of RPC auth flavours %d\n", + data->auth_flavourlen); + return -EINVAL; + +out_no_address: + dfprintk(MOUNT, "NFS4: mount program didn't pass remote address\n"); + return -EINVAL; } /* @@ -806,81 +1718,29 @@ static int nfs4_get_sb(struct file_system_type *fs_type, rpc_authflavor_t authflavour; struct nfs_fh mntfh; struct dentry *mntroot; - char *mntpath = NULL, *hostname = NULL, ip_addr[16]; - void *p; + char *mntpath = NULL, *hostname = NULL, *ip_addr = NULL; + int (*compare_super)(struct super_block *, void *) = nfs_compare_super; int error; - if (data == NULL) { - dprintk("%s: missing data argument\n", __FUNCTION__); - return -EINVAL; - } - if (data->version <= 0 || data->version > NFS4_MOUNT_VERSION) { - dprintk("%s: bad mount version\n", __FUNCTION__); - return -EINVAL; - } - - /* We now require that the mount process passes the remote address */ - if (data->host_addrlen != sizeof(addr)) - return -EINVAL; - - if (copy_from_user(&addr, data->host_addr, sizeof(addr))) - return -EFAULT; - - if (addr.sin_family != AF_INET || - addr.sin_addr.s_addr == INADDR_ANY - ) { - dprintk("%s: mount program didn't pass remote IP address!\n", - __FUNCTION__); - return -EINVAL; - } - /* RFC3530: The default port for NFS is 2049 */ - if (addr.sin_port == 0) - addr.sin_port = htons(NFS_PORT); - - /* Grab the authentication type */ - authflavour = RPC_AUTH_UNIX; - if (data->auth_flavourlen != 0) { - if (data->auth_flavourlen != 1) { - dprintk("%s: Invalid number of RPC auth flavours %d.\n", - __FUNCTION__, data->auth_flavourlen); - error = -EINVAL; - goto out_err_noserver; - } - - if (copy_from_user(&authflavour, data->auth_flavours, - sizeof(authflavour))) { - error = -EFAULT; - goto out_err_noserver; - } - } - - p = nfs_copy_user_string(NULL, &data->hostname, 256); - if (IS_ERR(p)) - goto out_err; - hostname = p; - - p = nfs_copy_user_string(NULL, &data->mnt_path, 1024); - if (IS_ERR(p)) - goto out_err; - mntpath = p; - - dprintk("MNTPATH: %s\n", mntpath); - - p = nfs_copy_user_string(ip_addr, &data->client_addr, - sizeof(ip_addr) - 1); - if (IS_ERR(p)) - goto out_err; + /* Validate the mount data */ + error = nfs4_validate_mount_data(&data, dev_name, &addr, &authflavour, + &hostname, &mntpath, &ip_addr); + if (error < 0) + goto out; /* Get a volume representation */ server = nfs4_create_server(data, hostname, &addr, mntpath, ip_addr, authflavour, &mntfh); if (IS_ERR(server)) { error = PTR_ERR(server); - goto out_err_noserver; + goto out; } + if (server->flags & NFS4_MOUNT_UNSHARED) + compare_super = NULL; + /* Get a superblock - note that we may end up sharing one that already exists */ - s = sget(fs_type, nfs_compare_super, nfs_set_super, server); + s = sget(fs_type, compare_super, nfs_set_super, server); if (IS_ERR(s)) { error = PTR_ERR(s); goto out_free; @@ -906,25 +1766,22 @@ static int nfs4_get_sb(struct file_system_type *fs_type, s->s_flags |= MS_ACTIVE; mnt->mnt_sb = s; mnt->mnt_root = mntroot; + error = 0; + +out: + kfree(ip_addr); kfree(mntpath); kfree(hostname); - return 0; - -out_err: - error = PTR_ERR(p); - goto out_err_noserver; + return error; out_free: nfs_free_server(server); -out_err_noserver: - kfree(mntpath); - kfree(hostname); - return error; + goto out; error_splat_super: up_write(&s->s_umount); deactivate_super(s); - goto out_err_noserver; + goto out; } static void nfs4_kill_super(struct super_block *sb) @@ -949,6 +1806,7 @@ static int nfs4_xdev_get_sb(struct file_system_type *fs_type, int flags, struct super_block *s; struct nfs_server *server; struct dentry *mntroot; + int (*compare_super)(struct super_block *, void *) = nfs_compare_super; int error; dprintk("--> nfs4_xdev_get_sb()\n"); @@ -960,8 +1818,11 @@ static int nfs4_xdev_get_sb(struct file_system_type *fs_type, int flags, goto out_err_noserver; } + if (server->flags & NFS4_MOUNT_UNSHARED) + compare_super = NULL; + /* Get a superblock - note that we may end up sharing one that already exists */ - s = sget(&nfs_fs_type, nfs_compare_super, nfs_set_super, server); + s = sget(&nfs_fs_type, compare_super, nfs_set_super, server); if (IS_ERR(s)) { error = PTR_ERR(s); goto out_err_nosb; @@ -1016,6 +1877,7 @@ static int nfs4_referral_get_sb(struct file_system_type *fs_type, int flags, struct nfs_server *server; struct dentry *mntroot; struct nfs_fh mntfh; + int (*compare_super)(struct super_block *, void *) = nfs_compare_super; int error; dprintk("--> nfs4_referral_get_sb()\n"); @@ -1027,8 +1889,11 @@ static int nfs4_referral_get_sb(struct file_system_type *fs_type, int flags, goto out_err_noserver; } + if (server->flags & NFS4_MOUNT_UNSHARED) + compare_super = NULL; + /* Get a superblock - note that we may end up sharing one that already exists */ - s = sget(&nfs_fs_type, nfs_compare_super, nfs_set_super, server); + s = sget(&nfs_fs_type, compare_super, nfs_set_super, server); if (IS_ERR(s)) { error = PTR_ERR(s); goto out_err_nosb; diff --git a/fs/nfs/write.c b/fs/nfs/write.c index af344a158e01..73ac992ece85 100644 --- a/fs/nfs/write.c +++ b/fs/nfs/write.c @@ -117,19 +117,19 @@ static struct nfs_page *nfs_page_find_request_locked(struct page *page) if (PagePrivate(page)) { req = (struct nfs_page *)page_private(page); if (req != NULL) - atomic_inc(&req->wb_count); + kref_get(&req->wb_kref); } return req; } static struct nfs_page *nfs_page_find_request(struct page *page) { + struct inode *inode = page->mapping->host; struct nfs_page *req = NULL; - spinlock_t *req_lock = &NFS_I(page->mapping->host)->req_lock; - spin_lock(req_lock); + spin_lock(&inode->i_lock); req = nfs_page_find_request_locked(page); - spin_unlock(req_lock); + spin_unlock(&inode->i_lock); return req; } @@ -191,8 +191,6 @@ static int nfs_writepage_setup(struct nfs_open_context *ctx, struct page *page, } /* Update file length */ nfs_grow_file(page, offset, count); - /* Set the PG_uptodate flag? */ - nfs_mark_uptodate(page, offset, count); nfs_unlock_request(req); return 0; } @@ -253,16 +251,16 @@ static void nfs_end_page_writeback(struct page *page) static int nfs_page_async_flush(struct nfs_pageio_descriptor *pgio, struct page *page) { + struct inode *inode = page->mapping->host; + struct nfs_inode *nfsi = NFS_I(inode); struct nfs_page *req; - struct nfs_inode *nfsi = NFS_I(page->mapping->host); - spinlock_t *req_lock = &nfsi->req_lock; int ret; - spin_lock(req_lock); + spin_lock(&inode->i_lock); for(;;) { req = nfs_page_find_request_locked(page); if (req == NULL) { - spin_unlock(req_lock); + spin_unlock(&inode->i_lock); return 1; } if (nfs_lock_request_dontget(req)) @@ -272,28 +270,28 @@ static int nfs_page_async_flush(struct nfs_pageio_descriptor *pgio, * succeed provided that someone hasn't already marked the * request as dirty (in which case we don't care). */ - spin_unlock(req_lock); + spin_unlock(&inode->i_lock); ret = nfs_wait_on_request(req); nfs_release_request(req); if (ret != 0) return ret; - spin_lock(req_lock); + spin_lock(&inode->i_lock); } if (test_bit(PG_NEED_COMMIT, &req->wb_flags)) { /* This request is marked for commit */ - spin_unlock(req_lock); + spin_unlock(&inode->i_lock); nfs_unlock_request(req); nfs_pageio_complete(pgio); return 1; } if (nfs_set_page_writeback(page) != 0) { - spin_unlock(req_lock); + spin_unlock(&inode->i_lock); BUG(); } radix_tree_tag_set(&nfsi->nfs_page_tree, req->wb_index, - NFS_PAGE_TAG_WRITEBACK); + NFS_PAGE_TAG_LOCKED); ret = test_bit(PG_NEED_FLUSH, &req->wb_flags); - spin_unlock(req_lock); + spin_unlock(&inode->i_lock); nfs_pageio_add_request(pgio, req); return ret; } @@ -400,7 +398,7 @@ static int nfs_inode_add_request(struct inode *inode, struct nfs_page *req) if (PageDirty(req->wb_page)) set_bit(PG_NEED_FLUSH, &req->wb_flags); nfsi->npages++; - atomic_inc(&req->wb_count); + kref_get(&req->wb_kref); return 0; } @@ -409,12 +407,12 @@ static int nfs_inode_add_request(struct inode *inode, struct nfs_page *req) */ static void nfs_inode_remove_request(struct nfs_page *req) { - struct inode *inode = req->wb_context->dentry->d_inode; + struct inode *inode = req->wb_context->path.dentry->d_inode; struct nfs_inode *nfsi = NFS_I(inode); BUG_ON (!NFS_WBACK_BUSY(req)); - spin_lock(&nfsi->req_lock); + spin_lock(&inode->i_lock); set_page_private(req->wb_page, 0); ClearPagePrivate(req->wb_page); radix_tree_delete(&nfsi->nfs_page_tree, req->wb_index); @@ -422,11 +420,11 @@ static void nfs_inode_remove_request(struct nfs_page *req) __set_page_dirty_nobuffers(req->wb_page); nfsi->npages--; if (!nfsi->npages) { - spin_unlock(&nfsi->req_lock); + spin_unlock(&inode->i_lock); nfs_end_data_update(inode); iput(inode); } else - spin_unlock(&nfsi->req_lock); + spin_unlock(&inode->i_lock); nfs_clear_request(req); nfs_release_request(req); } @@ -457,14 +455,16 @@ nfs_dirty_request(struct nfs_page *req) static void nfs_mark_request_commit(struct nfs_page *req) { - struct inode *inode = req->wb_context->dentry->d_inode; + struct inode *inode = req->wb_context->path.dentry->d_inode; struct nfs_inode *nfsi = NFS_I(inode); - spin_lock(&nfsi->req_lock); - nfs_list_add_request(req, &nfsi->commit); + spin_lock(&inode->i_lock); nfsi->ncommit++; set_bit(PG_NEED_COMMIT, &(req)->wb_flags); - spin_unlock(&nfsi->req_lock); + radix_tree_tag_set(&nfsi->nfs_page_tree, + req->wb_index, + NFS_PAGE_TAG_COMMIT); + spin_unlock(&inode->i_lock); inc_zone_page_state(req->wb_page, NR_UNSTABLE_NFS); __mark_inode_dirty(inode, I_DIRTY_DATASYNC); } @@ -526,18 +526,18 @@ static int nfs_wait_on_requests_locked(struct inode *inode, pgoff_t idx_start, u idx_end = idx_start + npages - 1; next = idx_start; - while (radix_tree_gang_lookup_tag(&nfsi->nfs_page_tree, (void **)&req, next, 1, NFS_PAGE_TAG_WRITEBACK)) { + while (radix_tree_gang_lookup_tag(&nfsi->nfs_page_tree, (void **)&req, next, 1, NFS_PAGE_TAG_LOCKED)) { if (req->wb_index > idx_end) break; next = req->wb_index + 1; BUG_ON(!NFS_WBACK_BUSY(req)); - atomic_inc(&req->wb_count); - spin_unlock(&nfsi->req_lock); + kref_get(&req->wb_kref); + spin_unlock(&inode->i_lock); error = nfs_wait_on_request(req); nfs_release_request(req); - spin_lock(&nfsi->req_lock); + spin_lock(&inode->i_lock); if (error < 0) return error; res++; @@ -577,10 +577,9 @@ nfs_scan_commit(struct inode *inode, struct list_head *dst, pgoff_t idx_start, u int res = 0; if (nfsi->ncommit != 0) { - res = nfs_scan_list(nfsi, &nfsi->commit, dst, idx_start, npages); + res = nfs_scan_list(nfsi, dst, idx_start, npages, + NFS_PAGE_TAG_COMMIT); nfsi->ncommit -= res; - if ((nfsi->ncommit == 0) != list_empty(&nfsi->commit)) - printk(KERN_ERR "NFS: desynchronized value of nfs_i.ncommit.\n"); } return res; } @@ -603,7 +602,6 @@ static struct nfs_page * nfs_update_request(struct nfs_open_context* ctx, { struct address_space *mapping = page->mapping; struct inode *inode = mapping->host; - struct nfs_inode *nfsi = NFS_I(inode); struct nfs_page *req, *new = NULL; pgoff_t rqend, end; @@ -613,13 +611,13 @@ static struct nfs_page * nfs_update_request(struct nfs_open_context* ctx, /* Loop over all inode entries and see if we find * A request for the page we wish to update */ - spin_lock(&nfsi->req_lock); + spin_lock(&inode->i_lock); req = nfs_page_find_request_locked(page); if (req) { if (!nfs_lock_request_dontget(req)) { int error; - spin_unlock(&nfsi->req_lock); + spin_unlock(&inode->i_lock); error = nfs_wait_on_request(req); nfs_release_request(req); if (error < 0) { @@ -629,7 +627,7 @@ static struct nfs_page * nfs_update_request(struct nfs_open_context* ctx, } continue; } - spin_unlock(&nfsi->req_lock); + spin_unlock(&inode->i_lock); if (new) nfs_release_request(new); break; @@ -640,14 +638,14 @@ static struct nfs_page * nfs_update_request(struct nfs_open_context* ctx, nfs_lock_request_dontget(new); error = nfs_inode_add_request(inode, new); if (error) { - spin_unlock(&nfsi->req_lock); + spin_unlock(&inode->i_lock); nfs_unlock_request(new); return ERR_PTR(error); } - spin_unlock(&nfsi->req_lock); + spin_unlock(&inode->i_lock); return new; } - spin_unlock(&nfsi->req_lock); + spin_unlock(&inode->i_lock); new = nfs_create_request(ctx, inode, page, offset, bytes); if (IS_ERR(new)) @@ -751,12 +749,17 @@ int nfs_updatepage(struct file *file, struct page *page, static void nfs_writepage_release(struct nfs_page *req) { - if (PageError(req->wb_page) || !nfs_reschedule_unstable_write(req)) { + if (PageError(req->wb_page)) { + nfs_end_page_writeback(req->wb_page); + nfs_inode_remove_request(req); + } else if (!nfs_reschedule_unstable_write(req)) { + /* Set the PG_uptodate flag */ + nfs_mark_uptodate(req->wb_page, req->wb_pgbase, req->wb_bytes); nfs_end_page_writeback(req->wb_page); nfs_inode_remove_request(req); } else nfs_end_page_writeback(req->wb_page); - nfs_clear_page_writeback(req); + nfs_clear_page_tag_locked(req); } static inline int flush_task_priority(int how) @@ -786,7 +789,7 @@ static void nfs_write_rpcsetup(struct nfs_page *req, * NB: take care not to mess about with data->commit et al. */ data->req = req; - data->inode = inode = req->wb_context->dentry->d_inode; + data->inode = inode = req->wb_context->path.dentry->d_inode; data->cred = req->wb_context->cred; data->args.fh = NFS_FH(inode); @@ -885,7 +888,7 @@ out_bad: } nfs_redirty_request(req); nfs_end_page_writeback(req->wb_page); - nfs_clear_page_writeback(req); + nfs_clear_page_tag_locked(req); return -ENOMEM; } @@ -928,7 +931,7 @@ static int nfs_flush_one(struct inode *inode, struct list_head *head, unsigned i nfs_list_remove_request(req); nfs_redirty_request(req); nfs_end_page_writeback(req->wb_page); - nfs_clear_page_writeback(req); + nfs_clear_page_tag_locked(req); } return -ENOMEM; } @@ -954,8 +957,8 @@ static void nfs_writeback_done_partial(struct rpc_task *task, void *calldata) struct page *page = req->wb_page; dprintk("NFS: write (%s/%Ld %d@%Ld)", - req->wb_context->dentry->d_inode->i_sb->s_id, - (long long)NFS_FILEID(req->wb_context->dentry->d_inode), + req->wb_context->path.dentry->d_inode->i_sb->s_id, + (long long)NFS_FILEID(req->wb_context->path.dentry->d_inode), req->wb_bytes, (long long)req_offset(req)); @@ -970,9 +973,9 @@ static void nfs_writeback_done_partial(struct rpc_task *task, void *calldata) } if (nfs_write_need_commit(data)) { - spinlock_t *req_lock = &NFS_I(page->mapping->host)->req_lock; + struct inode *inode = page->mapping->host; - spin_lock(req_lock); + spin_lock(&inode->i_lock); if (test_bit(PG_NEED_RESCHED, &req->wb_flags)) { /* Do nothing we need to resend the writes */ } else if (!test_and_set_bit(PG_NEED_COMMIT, &req->wb_flags)) { @@ -983,7 +986,7 @@ static void nfs_writeback_done_partial(struct rpc_task *task, void *calldata) clear_bit(PG_NEED_COMMIT, &req->wb_flags); dprintk(" server reboot detected\n"); } - spin_unlock(req_lock); + spin_unlock(&inode->i_lock); } else dprintk(" OK\n"); @@ -1020,8 +1023,8 @@ static void nfs_writeback_done_full(struct rpc_task *task, void *calldata) page = req->wb_page; dprintk("NFS: write (%s/%Ld %d@%Ld)", - req->wb_context->dentry->d_inode->i_sb->s_id, - (long long)NFS_FILEID(req->wb_context->dentry->d_inode), + req->wb_context->path.dentry->d_inode->i_sb->s_id, + (long long)NFS_FILEID(req->wb_context->path.dentry->d_inode), req->wb_bytes, (long long)req_offset(req)); @@ -1039,12 +1042,14 @@ static void nfs_writeback_done_full(struct rpc_task *task, void *calldata) dprintk(" marked for commit\n"); goto next; } + /* Set the PG_uptodate flag? */ + nfs_mark_uptodate(page, req->wb_pgbase, req->wb_bytes); dprintk(" OK\n"); remove_request: nfs_end_page_writeback(page); nfs_inode_remove_request(req); next: - nfs_clear_page_writeback(req); + nfs_clear_page_tag_locked(req); } } @@ -1157,7 +1162,7 @@ static void nfs_commit_rpcsetup(struct list_head *head, list_splice_init(head, &data->pages); first = nfs_list_entry(data->pages.next); - inode = first->wb_context->dentry->d_inode; + inode = first->wb_context->path.dentry->d_inode; data->inode = inode; data->cred = first->wb_context->cred; @@ -1207,7 +1212,7 @@ nfs_commit_list(struct inode *inode, struct list_head *head, int how) nfs_list_remove_request(req); nfs_mark_request_commit(req); dec_zone_page_state(req->wb_page, NR_UNSTABLE_NFS); - nfs_clear_page_writeback(req); + nfs_clear_page_tag_locked(req); } return -ENOMEM; } @@ -1234,8 +1239,8 @@ static void nfs_commit_done(struct rpc_task *task, void *calldata) dec_zone_page_state(req->wb_page, NR_UNSTABLE_NFS); dprintk("NFS: commit (%s/%Ld %d@%Ld)", - req->wb_context->dentry->d_inode->i_sb->s_id, - (long long)NFS_FILEID(req->wb_context->dentry->d_inode), + req->wb_context->path.dentry->d_inode->i_sb->s_id, + (long long)NFS_FILEID(req->wb_context->path.dentry->d_inode), req->wb_bytes, (long long)req_offset(req)); if (task->tk_status < 0) { @@ -1249,6 +1254,9 @@ static void nfs_commit_done(struct rpc_task *task, void *calldata) * returned by the server against all stored verfs. */ if (!memcmp(req->wb_verf.verifier, data->verf.verifier, sizeof(data->verf.verifier))) { /* We have a match */ + /* Set the PG_uptodate flag */ + nfs_mark_uptodate(req->wb_page, req->wb_pgbase, + req->wb_bytes); nfs_inode_remove_request(req); dprintk(" OK\n"); goto next; @@ -1257,7 +1265,7 @@ static void nfs_commit_done(struct rpc_task *task, void *calldata) dprintk(" mismatch\n"); nfs_redirty_request(req); next: - nfs_clear_page_writeback(req); + nfs_clear_page_tag_locked(req); } } @@ -1268,13 +1276,12 @@ static const struct rpc_call_ops nfs_commit_ops = { int nfs_commit_inode(struct inode *inode, int how) { - struct nfs_inode *nfsi = NFS_I(inode); LIST_HEAD(head); int res; - spin_lock(&nfsi->req_lock); + spin_lock(&inode->i_lock); res = nfs_scan_commit(inode, &head, 0, 0); - spin_unlock(&nfsi->req_lock); + spin_unlock(&inode->i_lock); if (res) { int error = nfs_commit_list(inode, &head, how); if (error < 0) @@ -1292,7 +1299,6 @@ static inline int nfs_commit_list(struct inode *inode, struct list_head *head, i long nfs_sync_mapping_wait(struct address_space *mapping, struct writeback_control *wbc, int how) { struct inode *inode = mapping->host; - struct nfs_inode *nfsi = NFS_I(inode); pgoff_t idx_start, idx_end; unsigned int npages = 0; LIST_HEAD(head); @@ -1314,7 +1320,7 @@ long nfs_sync_mapping_wait(struct address_space *mapping, struct writeback_contr } } how &= ~FLUSH_NOCOMMIT; - spin_lock(&nfsi->req_lock); + spin_lock(&inode->i_lock); do { ret = nfs_wait_on_requests_locked(inode, idx_start, npages); if (ret != 0) @@ -1325,18 +1331,19 @@ long nfs_sync_mapping_wait(struct address_space *mapping, struct writeback_contr if (pages == 0) break; if (how & FLUSH_INVALIDATE) { - spin_unlock(&nfsi->req_lock); + spin_unlock(&inode->i_lock); nfs_cancel_commit_list(&head); ret = pages; - spin_lock(&nfsi->req_lock); + spin_lock(&inode->i_lock); continue; } pages += nfs_scan_commit(inode, &head, 0, 0); - spin_unlock(&nfsi->req_lock); + spin_unlock(&inode->i_lock); ret = nfs_commit_list(inode, &head, how); - spin_lock(&nfsi->req_lock); + spin_lock(&inode->i_lock); + } while (ret >= 0); - spin_unlock(&nfsi->req_lock); + spin_unlock(&inode->i_lock); return ret; } @@ -1430,7 +1437,6 @@ int nfs_set_page_dirty(struct page *page) { struct address_space *mapping = page->mapping; struct inode *inode; - spinlock_t *req_lock; struct nfs_page *req; int ret; @@ -1439,18 +1445,17 @@ int nfs_set_page_dirty(struct page *page) inode = mapping->host; if (!inode) goto out_raced; - req_lock = &NFS_I(inode)->req_lock; - spin_lock(req_lock); + spin_lock(&inode->i_lock); req = nfs_page_find_request_locked(page); if (req != NULL) { /* Mark any existing write requests for flushing */ ret = !test_and_set_bit(PG_NEED_FLUSH, &req->wb_flags); - spin_unlock(req_lock); + spin_unlock(&inode->i_lock); nfs_release_request(req); return ret; } ret = __set_page_dirty_nobuffers(page); - spin_unlock(req_lock); + spin_unlock(&inode->i_lock); return ret; out_raced: return !TestSetPageDirty(page); diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c index 864090edc28b..5443c52b57aa 100644 --- a/fs/nfsd/nfs4callback.c +++ b/fs/nfsd/nfs4callback.c @@ -394,7 +394,6 @@ nfsd4_probe_callback(struct nfs4_client *clp) .rpc_proc = &nfs4_cb_procedures[NFSPROC4_CLNT_CB_NULL], .rpc_argp = clp, }; - char clientname[16]; int status; if (atomic_read(&cb->cb_set)) @@ -417,11 +416,6 @@ nfsd4_probe_callback(struct nfs4_client *clp) memset(program->stats, 0, sizeof(cb->cb_stat)); program->stats->program = program; - /* Just here to make some printk's more useful: */ - snprintf(clientname, sizeof(clientname), - "%u.%u.%u.%u", NIPQUAD(addr.sin_addr)); - args.servername = clientname; - /* Create RPC client */ cb->cb_client = rpc_create(&args); if (IS_ERR(cb->cb_client)) { @@ -429,29 +423,23 @@ nfsd4_probe_callback(struct nfs4_client *clp) goto out_err; } - /* Kick rpciod, put the call on the wire. */ - if (rpciod_up() != 0) - goto out_clnt; - /* the task holds a reference to the nfs4_client struct */ atomic_inc(&clp->cl_count); msg.rpc_cred = nfsd4_lookupcred(clp,0); if (IS_ERR(msg.rpc_cred)) - goto out_rpciod; + goto out_release_clp; status = rpc_call_async(cb->cb_client, &msg, RPC_TASK_ASYNC, &nfs4_cb_null_ops, NULL); put_rpccred(msg.rpc_cred); if (status != 0) { dprintk("NFSD: asynchronous NFSPROC4_CB_NULL failed!\n"); - goto out_rpciod; + goto out_release_clp; } return; -out_rpciod: +out_release_clp: atomic_dec(&clp->cl_count); - rpciod_down(); -out_clnt: rpc_shutdown_client(cb->cb_client); out_err: cb->cb_client = NULL; diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 3cc8ce422ab1..8c52913d7cb6 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -378,7 +378,6 @@ shutdown_callback_client(struct nfs4_client *clp) if (clnt) { clp->cl_callback.cb_client = NULL; rpc_shutdown_client(clnt); - rpciod_down(); } } diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c index 8604e35bd48e..945b1cedde2b 100644 --- a/fs/nfsd/vfs.c +++ b/fs/nfsd/vfs.c @@ -879,6 +879,7 @@ nfsd_vfs_read(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file, .u.data = rqstp, }; + rqstp->rq_resused = 1; host_err = splice_direct_to_actor(file, &sd, nfsd_direct_splice_actor); } else { oldfs = get_fs(); diff --git a/fs/ocfs2/cluster/masklog.c b/fs/ocfs2/cluster/masklog.c index 2b205f5d5790..e9e042b93dbf 100644 --- a/fs/ocfs2/cluster/masklog.c +++ b/fs/ocfs2/cluster/masklog.c @@ -74,7 +74,6 @@ struct mlog_attribute { #define define_mask(_name) { \ .attr = { \ .name = #_name, \ - .owner = THIS_MODULE, \ .mode = S_IRUGO | S_IWUSR, \ }, \ .mask = ML_##_name, \ diff --git a/fs/partitions/check.c b/fs/partitions/check.c index 9a3a058f3553..98e0b85a9bb2 100644 --- a/fs/partitions/check.c +++ b/fs/partitions/check.c @@ -397,7 +397,6 @@ void add_partition(struct gendisk *disk, int part, sector_t start, sector_t len, static struct attribute addpartattr = { .name = "whole_disk", .mode = S_IRUSR | S_IRGRP | S_IROTH, - .owner = THIS_MODULE, }; sysfs_create_file(&p->kobj, &addpartattr); diff --git a/fs/splice.c b/fs/splice.c index ed2ce995475c..6c9828651e6f 100644 --- a/fs/splice.c +++ b/fs/splice.c @@ -28,6 +28,7 @@ #include <linux/module.h> #include <linux/syscalls.h> #include <linux/uio.h> +#include <linux/security.h> /* * Attempt to steal a page from a pipe buffer. This should perhaps go into @@ -491,7 +492,7 @@ ssize_t generic_file_splice_read(struct file *in, loff_t *ppos, ret = 0; spliced = 0; - while (len) { + while (len && !spliced) { ret = __generic_file_splice_read(in, ppos, pipe, len, flags); if (ret < 0) @@ -961,6 +962,10 @@ static long do_splice_from(struct pipe_inode_info *pipe, struct file *out, if (unlikely(ret < 0)) return ret; + ret = security_file_permission(out, MAY_WRITE); + if (unlikely(ret < 0)) + return ret; + return out->f_op->splice_write(pipe, out, ppos, len, flags); } @@ -983,6 +988,10 @@ static long do_splice_to(struct file *in, loff_t *ppos, if (unlikely(ret < 0)) return ret; + ret = security_file_permission(in, MAY_READ); + if (unlikely(ret < 0)) + return ret; + return in->f_op->splice_read(in, ppos, pipe, len, flags); } @@ -1051,15 +1060,10 @@ ssize_t splice_direct_to_actor(struct file *in, struct splice_desc *sd, sd->flags &= ~SPLICE_F_NONBLOCK; while (len) { - size_t read_len, max_read_len; - - /* - * Do at most PIPE_BUFFERS pages worth of transfer: - */ - max_read_len = min(len, (size_t)(PIPE_BUFFERS*PAGE_SIZE)); + size_t read_len; - ret = do_splice_to(in, &sd->pos, pipe, max_read_len, flags); - if (unlikely(ret < 0)) + ret = do_splice_to(in, &sd->pos, pipe, len, flags); + if (unlikely(ret <= 0)) goto out_release; read_len = ret; @@ -1071,26 +1075,17 @@ ssize_t splice_direct_to_actor(struct file *in, struct splice_desc *sd, * could get stuck data in the internal pipe: */ ret = actor(pipe, sd); - if (unlikely(ret < 0)) + if (unlikely(ret <= 0)) goto out_release; bytes += ret; len -= ret; - /* - * In nonblocking mode, if we got back a short read then - * that was due to either an IO error or due to the - * pagecache entry not being there. In the IO error case - * the _next_ splice attempt will produce a clean IO error - * return value (not a short read), so in both cases it's - * correct to break out of the loop here: - */ - if ((flags & SPLICE_F_NONBLOCK) && (read_len < max_read_len)) - break; + if (ret < read_len) + goto out_release; } pipe->nrbufs = pipe->curbuf = 0; - return bytes; out_release: @@ -1152,10 +1147,12 @@ long do_splice_direct(struct file *in, loff_t *ppos, struct file *out, .pos = *ppos, .u.file = out, }; - size_t ret; + long ret; ret = splice_direct_to_actor(in, &sd, direct_splice_actor); - *ppos = sd.pos; + if (ret > 0) + *ppos += ret; + return ret; } diff --git a/fs/sysfs/bin.c b/fs/sysfs/bin.c index d3b9f5f07db1..135353f8a296 100644 --- a/fs/sysfs/bin.c +++ b/fs/sysfs/bin.c @@ -20,29 +20,41 @@ #include "sysfs.h" +struct bin_buffer { + struct mutex mutex; + void *buffer; + int mmapped; +}; + static int fill_read(struct dentry *dentry, char *buffer, loff_t off, size_t count) { - struct bin_attribute * attr = to_bin_attr(dentry); - struct kobject * kobj = to_kobj(dentry->d_parent); + struct sysfs_dirent *attr_sd = dentry->d_fsdata; + struct bin_attribute *attr = attr_sd->s_elem.bin_attr.bin_attr; + struct kobject *kobj = attr_sd->s_parent->s_elem.dir.kobj; + int rc; + + /* need attr_sd for attr, its parent for kobj */ + if (!sysfs_get_active_two(attr_sd)) + return -ENODEV; - if (!attr->read) - return -EIO; + rc = -EIO; + if (attr->read) + rc = attr->read(kobj, attr, buffer, off, count); - return attr->read(kobj, buffer, off, count); + sysfs_put_active_two(attr_sd); + + return rc; } static ssize_t -read(struct file * file, char __user * userbuf, size_t count, loff_t * off) +read(struct file *file, char __user *userbuf, size_t bytes, loff_t *off) { - char *buffer = file->private_data; + struct bin_buffer *bb = file->private_data; struct dentry *dentry = file->f_path.dentry; int size = dentry->d_inode->i_size; loff_t offs = *off; - int ret; - - if (count > PAGE_SIZE) - count = PAGE_SIZE; + int count = min_t(size_t, bytes, PAGE_SIZE); if (size) { if (offs > size) @@ -51,43 +63,56 @@ read(struct file * file, char __user * userbuf, size_t count, loff_t * off) count = size - offs; } - ret = fill_read(dentry, buffer, offs, count); - if (ret < 0) - return ret; - count = ret; + mutex_lock(&bb->mutex); + + count = fill_read(dentry, bb->buffer, offs, count); + if (count < 0) + goto out_unlock; - if (copy_to_user(userbuf, buffer, count)) - return -EFAULT; + if (copy_to_user(userbuf, bb->buffer, count)) { + count = -EFAULT; + goto out_unlock; + } - pr_debug("offs = %lld, *off = %lld, count = %zd\n", offs, *off, count); + pr_debug("offs = %lld, *off = %lld, count = %d\n", offs, *off, count); *off = offs + count; + out_unlock: + mutex_unlock(&bb->mutex); return count; } static int flush_write(struct dentry *dentry, char *buffer, loff_t offset, size_t count) { - struct bin_attribute *attr = to_bin_attr(dentry); - struct kobject *kobj = to_kobj(dentry->d_parent); + struct sysfs_dirent *attr_sd = dentry->d_fsdata; + struct bin_attribute *attr = attr_sd->s_elem.bin_attr.bin_attr; + struct kobject *kobj = attr_sd->s_parent->s_elem.dir.kobj; + int rc; + + /* need attr_sd for attr, its parent for kobj */ + if (!sysfs_get_active_two(attr_sd)) + return -ENODEV; + + rc = -EIO; + if (attr->write) + rc = attr->write(kobj, attr, buffer, offset, count); - if (!attr->write) - return -EIO; + sysfs_put_active_two(attr_sd); - return attr->write(kobj, buffer, offset, count); + return rc; } -static ssize_t write(struct file * file, const char __user * userbuf, - size_t count, loff_t * off) +static ssize_t write(struct file *file, const char __user *userbuf, + size_t bytes, loff_t *off) { - char *buffer = file->private_data; + struct bin_buffer *bb = file->private_data; struct dentry *dentry = file->f_path.dentry; int size = dentry->d_inode->i_size; loff_t offs = *off; + int count = min_t(size_t, bytes, PAGE_SIZE); - if (count > PAGE_SIZE) - count = PAGE_SIZE; if (size) { if (offs > size) return 0; @@ -95,72 +120,100 @@ static ssize_t write(struct file * file, const char __user * userbuf, count = size - offs; } - if (copy_from_user(buffer, userbuf, count)) - return -EFAULT; + mutex_lock(&bb->mutex); - count = flush_write(dentry, buffer, offs, count); + if (copy_from_user(bb->buffer, userbuf, count)) { + count = -EFAULT; + goto out_unlock; + } + + count = flush_write(dentry, bb->buffer, offs, count); if (count > 0) *off = offs + count; + + out_unlock: + mutex_unlock(&bb->mutex); return count; } static int mmap(struct file *file, struct vm_area_struct *vma) { - struct dentry *dentry = file->f_path.dentry; - struct bin_attribute *attr = to_bin_attr(dentry); - struct kobject *kobj = to_kobj(dentry->d_parent); + struct bin_buffer *bb = file->private_data; + struct sysfs_dirent *attr_sd = file->f_path.dentry->d_fsdata; + struct bin_attribute *attr = attr_sd->s_elem.bin_attr.bin_attr; + struct kobject *kobj = attr_sd->s_parent->s_elem.dir.kobj; + int rc; + + mutex_lock(&bb->mutex); + + /* need attr_sd for attr, its parent for kobj */ + if (!sysfs_get_active_two(attr_sd)) + return -ENODEV; - if (!attr->mmap) - return -EINVAL; + rc = -EINVAL; + if (attr->mmap) + rc = attr->mmap(kobj, attr, vma); - return attr->mmap(kobj, attr, vma); + if (rc == 0 && !bb->mmapped) + bb->mmapped = 1; + else + sysfs_put_active_two(attr_sd); + + mutex_unlock(&bb->mutex); + + return rc; } static int open(struct inode * inode, struct file * file) { - struct kobject *kobj = sysfs_get_kobject(file->f_path.dentry->d_parent); - struct bin_attribute * attr = to_bin_attr(file->f_path.dentry); - int error = -EINVAL; - - if (!kobj || !attr) - goto Done; + struct sysfs_dirent *attr_sd = file->f_path.dentry->d_fsdata; + struct bin_attribute *attr = attr_sd->s_elem.bin_attr.bin_attr; + struct bin_buffer *bb = NULL; + int error; - /* Grab the module reference for this attribute if we have one */ - error = -ENODEV; - if (!try_module_get(attr->attr.owner)) - goto Done; + /* need attr_sd for attr */ + if (!sysfs_get_active(attr_sd)) + return -ENODEV; error = -EACCES; if ((file->f_mode & FMODE_WRITE) && !(attr->write || attr->mmap)) - goto Error; + goto err_out; if ((file->f_mode & FMODE_READ) && !(attr->read || attr->mmap)) - goto Error; + goto err_out; error = -ENOMEM; - file->private_data = kmalloc(PAGE_SIZE, GFP_KERNEL); - if (!file->private_data) - goto Error; - - error = 0; - goto Done; - - Error: - module_put(attr->attr.owner); - Done: - if (error) - kobject_put(kobj); + bb = kzalloc(sizeof(*bb), GFP_KERNEL); + if (!bb) + goto err_out; + + bb->buffer = kmalloc(PAGE_SIZE, GFP_KERNEL); + if (!bb->buffer) + goto err_out; + + mutex_init(&bb->mutex); + file->private_data = bb; + + /* open succeeded, put active reference and pin attr_sd */ + sysfs_put_active(attr_sd); + sysfs_get(attr_sd); + return 0; + + err_out: + sysfs_put_active(attr_sd); + kfree(bb); return error; } static int release(struct inode * inode, struct file * file) { - struct kobject * kobj = to_kobj(file->f_path.dentry->d_parent); - struct bin_attribute * attr = to_bin_attr(file->f_path.dentry); - u8 * buffer = file->private_data; - - kobject_put(kobj); - module_put(attr->attr.owner); - kfree(buffer); + struct sysfs_dirent *attr_sd = file->f_path.dentry->d_fsdata; + struct bin_buffer *bb = file->private_data; + + if (bb->mmapped) + sysfs_put_active_two(attr_sd); + sysfs_put(attr_sd); + kfree(bb->buffer); + kfree(bb); return 0; } @@ -181,9 +234,9 @@ const struct file_operations bin_fops = { int sysfs_create_bin_file(struct kobject * kobj, struct bin_attribute * attr) { - BUG_ON(!kobj || !kobj->dentry || !attr); + BUG_ON(!kobj || !kobj->sd || !attr); - return sysfs_add_file(kobj->dentry, &attr->attr, SYSFS_KOBJ_BIN_ATTR); + return sysfs_add_file(kobj->sd, &attr->attr, SYSFS_KOBJ_BIN_ATTR); } @@ -195,7 +248,7 @@ int sysfs_create_bin_file(struct kobject * kobj, struct bin_attribute * attr) void sysfs_remove_bin_file(struct kobject * kobj, struct bin_attribute * attr) { - if (sysfs_hash_and_remove(kobj->dentry, attr->attr.name) < 0) { + if (sysfs_hash_and_remove(kobj->sd, attr->attr.name) < 0) { printk(KERN_ERR "%s: " "bad dentry or inode or no such file: \"%s\"\n", __FUNCTION__, attr->attr.name); diff --git a/fs/sysfs/dir.c b/fs/sysfs/dir.c index c4342a019972..aee966c44aac 100644 --- a/fs/sysfs/dir.c +++ b/fs/sysfs/dir.c @@ -9,21 +9,337 @@ #include <linux/module.h> #include <linux/kobject.h> #include <linux/namei.h> +#include <linux/idr.h> +#include <linux/completion.h> #include <asm/semaphore.h> #include "sysfs.h" -DECLARE_RWSEM(sysfs_rename_sem); -spinlock_t sysfs_lock = SPIN_LOCK_UNLOCKED; +DEFINE_MUTEX(sysfs_mutex); +spinlock_t sysfs_assoc_lock = SPIN_LOCK_UNLOCKED; + +static spinlock_t sysfs_ino_lock = SPIN_LOCK_UNLOCKED; +static DEFINE_IDA(sysfs_ino_ida); + +/** + * sysfs_link_sibling - link sysfs_dirent into sibling list + * @sd: sysfs_dirent of interest + * + * Link @sd into its sibling list which starts from + * sd->s_parent->s_children. + * + * Locking: + * mutex_lock(sysfs_mutex) + */ +void sysfs_link_sibling(struct sysfs_dirent *sd) +{ + struct sysfs_dirent *parent_sd = sd->s_parent; + + BUG_ON(sd->s_sibling); + sd->s_sibling = parent_sd->s_children; + parent_sd->s_children = sd; +} + +/** + * sysfs_unlink_sibling - unlink sysfs_dirent from sibling list + * @sd: sysfs_dirent of interest + * + * Unlink @sd from its sibling list which starts from + * sd->s_parent->s_children. + * + * Locking: + * mutex_lock(sysfs_mutex) + */ +void sysfs_unlink_sibling(struct sysfs_dirent *sd) +{ + struct sysfs_dirent **pos; + + for (pos = &sd->s_parent->s_children; *pos; pos = &(*pos)->s_sibling) { + if (*pos == sd) { + *pos = sd->s_sibling; + sd->s_sibling = NULL; + break; + } + } +} + +/** + * sysfs_get_dentry - get dentry for the given sysfs_dirent + * @sd: sysfs_dirent of interest + * + * Get dentry for @sd. Dentry is looked up if currently not + * present. This function climbs sysfs_dirent tree till it + * reaches a sysfs_dirent with valid dentry attached and descends + * down from there looking up dentry for each step. + * + * LOCKING: + * Kernel thread context (may sleep) + * + * RETURNS: + * Pointer to found dentry on success, ERR_PTR() value on error. + */ +struct dentry *sysfs_get_dentry(struct sysfs_dirent *sd) +{ + struct sysfs_dirent *cur; + struct dentry *parent_dentry, *dentry; + int i, depth; + + /* Find the first parent which has valid s_dentry and get the + * dentry. + */ + mutex_lock(&sysfs_mutex); + restart0: + spin_lock(&sysfs_assoc_lock); + restart1: + spin_lock(&dcache_lock); + + dentry = NULL; + depth = 0; + cur = sd; + while (!cur->s_dentry || !cur->s_dentry->d_inode) { + if (cur->s_flags & SYSFS_FLAG_REMOVED) { + dentry = ERR_PTR(-ENOENT); + depth = 0; + break; + } + cur = cur->s_parent; + depth++; + } + if (!IS_ERR(dentry)) + dentry = dget_locked(cur->s_dentry); + + spin_unlock(&dcache_lock); + spin_unlock(&sysfs_assoc_lock); + + /* from the found dentry, look up depth times */ + while (depth--) { + /* find and get depth'th ancestor */ + for (cur = sd, i = 0; cur && i < depth; i++) + cur = cur->s_parent; + + /* This can happen if tree structure was modified due + * to move/rename. Restart. + */ + if (i != depth) { + dput(dentry); + goto restart0; + } + + sysfs_get(cur); + + mutex_unlock(&sysfs_mutex); + + /* look it up */ + parent_dentry = dentry; + dentry = lookup_one_len_kern(cur->s_name, parent_dentry, + strlen(cur->s_name)); + dput(parent_dentry); + + if (IS_ERR(dentry)) { + sysfs_put(cur); + return dentry; + } + + mutex_lock(&sysfs_mutex); + spin_lock(&sysfs_assoc_lock); + + /* This, again, can happen if tree structure has + * changed and we looked up the wrong thing. Restart. + */ + if (cur->s_dentry != dentry) { + dput(dentry); + sysfs_put(cur); + goto restart1; + } + + spin_unlock(&sysfs_assoc_lock); + + sysfs_put(cur); + } + + mutex_unlock(&sysfs_mutex); + return dentry; +} + +/** + * sysfs_get_active - get an active reference to sysfs_dirent + * @sd: sysfs_dirent to get an active reference to + * + * Get an active reference of @sd. This function is noop if @sd + * is NULL. + * + * RETURNS: + * Pointer to @sd on success, NULL on failure. + */ +struct sysfs_dirent *sysfs_get_active(struct sysfs_dirent *sd) +{ + if (unlikely(!sd)) + return NULL; + + while (1) { + int v, t; + + v = atomic_read(&sd->s_active); + if (unlikely(v < 0)) + return NULL; + + t = atomic_cmpxchg(&sd->s_active, v, v + 1); + if (likely(t == v)) + return sd; + if (t < 0) + return NULL; + + cpu_relax(); + } +} + +/** + * sysfs_put_active - put an active reference to sysfs_dirent + * @sd: sysfs_dirent to put an active reference to + * + * Put an active reference to @sd. This function is noop if @sd + * is NULL. + */ +void sysfs_put_active(struct sysfs_dirent *sd) +{ + struct completion *cmpl; + int v; + + if (unlikely(!sd)) + return; + + v = atomic_dec_return(&sd->s_active); + if (likely(v != SD_DEACTIVATED_BIAS)) + return; + + /* atomic_dec_return() is a mb(), we'll always see the updated + * sd->s_sibling. + */ + cmpl = (void *)sd->s_sibling; + complete(cmpl); +} + +/** + * sysfs_get_active_two - get active references to sysfs_dirent and parent + * @sd: sysfs_dirent of interest + * + * Get active reference to @sd and its parent. Parent's active + * reference is grabbed first. This function is noop if @sd is + * NULL. + * + * RETURNS: + * Pointer to @sd on success, NULL on failure. + */ +struct sysfs_dirent *sysfs_get_active_two(struct sysfs_dirent *sd) +{ + if (sd) { + if (sd->s_parent && unlikely(!sysfs_get_active(sd->s_parent))) + return NULL; + if (unlikely(!sysfs_get_active(sd))) { + sysfs_put_active(sd->s_parent); + return NULL; + } + } + return sd; +} + +/** + * sysfs_put_active_two - put active references to sysfs_dirent and parent + * @sd: sysfs_dirent of interest + * + * Put active references to @sd and its parent. This function is + * noop if @sd is NULL. + */ +void sysfs_put_active_two(struct sysfs_dirent *sd) +{ + if (sd) { + sysfs_put_active(sd); + sysfs_put_active(sd->s_parent); + } +} + +/** + * sysfs_deactivate - deactivate sysfs_dirent + * @sd: sysfs_dirent to deactivate + * + * Deny new active references and drain existing ones. + */ +static void sysfs_deactivate(struct sysfs_dirent *sd) +{ + DECLARE_COMPLETION_ONSTACK(wait); + int v; + + BUG_ON(sd->s_sibling || !(sd->s_flags & SYSFS_FLAG_REMOVED)); + sd->s_sibling = (void *)&wait; + + /* atomic_add_return() is a mb(), put_active() will always see + * the updated sd->s_sibling. + */ + v = atomic_add_return(SD_DEACTIVATED_BIAS, &sd->s_active); + + if (v != SD_DEACTIVATED_BIAS) + wait_for_completion(&wait); + + sd->s_sibling = NULL; +} + +static int sysfs_alloc_ino(ino_t *pino) +{ + int ino, rc; + + retry: + spin_lock(&sysfs_ino_lock); + rc = ida_get_new_above(&sysfs_ino_ida, 2, &ino); + spin_unlock(&sysfs_ino_lock); + + if (rc == -EAGAIN) { + if (ida_pre_get(&sysfs_ino_ida, GFP_KERNEL)) + goto retry; + rc = -ENOMEM; + } + + *pino = ino; + return rc; +} + +static void sysfs_free_ino(ino_t ino) +{ + spin_lock(&sysfs_ino_lock); + ida_remove(&sysfs_ino_ida, ino); + spin_unlock(&sysfs_ino_lock); +} + +void release_sysfs_dirent(struct sysfs_dirent * sd) +{ + struct sysfs_dirent *parent_sd; + + repeat: + /* Moving/renaming is always done while holding reference. + * sd->s_parent won't change beneath us. + */ + parent_sd = sd->s_parent; + + if (sysfs_type(sd) == SYSFS_KOBJ_LINK) + sysfs_put(sd->s_elem.symlink.target_sd); + if (sysfs_type(sd) & SYSFS_COPY_NAME) + kfree(sd->s_name); + kfree(sd->s_iattr); + sysfs_free_ino(sd->s_ino); + kmem_cache_free(sysfs_dir_cachep, sd); + + sd = parent_sd; + if (sd && atomic_dec_and_test(&sd->s_count)) + goto repeat; +} static void sysfs_d_iput(struct dentry * dentry, struct inode * inode) { struct sysfs_dirent * sd = dentry->d_fsdata; if (sd) { - /* sd->s_dentry is protected with sysfs_lock. This - * allows sysfs_drop_dentry() to dereference it. + /* sd->s_dentry is protected with sysfs_assoc_lock. + * This allows sysfs_drop_dentry() to dereference it. */ - spin_lock(&sysfs_lock); + spin_lock(&sysfs_assoc_lock); /* The dentry might have been deleted or another * lookup could have happened updating sd->s_dentry to @@ -32,7 +348,7 @@ static void sysfs_d_iput(struct dentry * dentry, struct inode * inode) */ if (sd->s_dentry == dentry) sd->s_dentry = NULL; - spin_unlock(&sysfs_lock); + spin_unlock(&sysfs_assoc_lock); sysfs_put(sd); } iput(inode); @@ -42,260 +358,402 @@ static struct dentry_operations sysfs_dentry_ops = { .d_iput = sysfs_d_iput, }; -static unsigned int sysfs_inode_counter; -ino_t sysfs_get_inum(void) +struct sysfs_dirent *sysfs_new_dirent(const char *name, umode_t mode, int type) { - if (unlikely(sysfs_inode_counter < 3)) - sysfs_inode_counter = 3; - return sysfs_inode_counter++; -} + char *dup_name = NULL; + struct sysfs_dirent *sd = NULL; -/* - * Allocates a new sysfs_dirent and links it to the parent sysfs_dirent - */ -static struct sysfs_dirent * __sysfs_new_dirent(void * element) -{ - struct sysfs_dirent * sd; + if (type & SYSFS_COPY_NAME) { + name = dup_name = kstrdup(name, GFP_KERNEL); + if (!name) + goto err_out; + } sd = kmem_cache_zalloc(sysfs_dir_cachep, GFP_KERNEL); if (!sd) - return NULL; + goto err_out; + + if (sysfs_alloc_ino(&sd->s_ino)) + goto err_out; - sd->s_ino = sysfs_get_inum(); atomic_set(&sd->s_count, 1); + atomic_set(&sd->s_active, 0); atomic_set(&sd->s_event, 1); - INIT_LIST_HEAD(&sd->s_children); - INIT_LIST_HEAD(&sd->s_sibling); - sd->s_element = element; + + sd->s_name = name; + sd->s_mode = mode; + sd->s_flags = type; return sd; + + err_out: + kfree(dup_name); + kmem_cache_free(sysfs_dir_cachep, sd); + return NULL; } -static void __sysfs_list_dirent(struct sysfs_dirent *parent_sd, - struct sysfs_dirent *sd) +/** + * sysfs_attach_dentry - associate sysfs_dirent with dentry + * @sd: target sysfs_dirent + * @dentry: dentry to associate + * + * Associate @sd with @dentry. This is protected by + * sysfs_assoc_lock to avoid race with sysfs_d_iput(). + * + * LOCKING: + * mutex_lock(sysfs_mutex) + */ +static void sysfs_attach_dentry(struct sysfs_dirent *sd, struct dentry *dentry) { - if (sd) - list_add(&sd->s_sibling, &parent_sd->s_children); + dentry->d_op = &sysfs_dentry_ops; + dentry->d_fsdata = sysfs_get(sd); + + /* protect sd->s_dentry against sysfs_d_iput */ + spin_lock(&sysfs_assoc_lock); + sd->s_dentry = dentry; + spin_unlock(&sysfs_assoc_lock); + + d_rehash(dentry); } -static struct sysfs_dirent * sysfs_new_dirent(struct sysfs_dirent *parent_sd, - void * element) +static int sysfs_ilookup_test(struct inode *inode, void *arg) { - struct sysfs_dirent *sd; - sd = __sysfs_new_dirent(element); - __sysfs_list_dirent(parent_sd, sd); - return sd; + struct sysfs_dirent *sd = arg; + return inode->i_ino == sd->s_ino; } -/* +/** + * sysfs_addrm_start - prepare for sysfs_dirent add/remove + * @acxt: pointer to sysfs_addrm_cxt to be used + * @parent_sd: parent sysfs_dirent * - * Return -EEXIST if there is already a sysfs element with the same name for - * the same parent. + * This function is called when the caller is about to add or + * remove sysfs_dirent under @parent_sd. This function acquires + * sysfs_mutex, grabs inode for @parent_sd if available and lock + * i_mutex of it. @acxt is used to keep and pass context to + * other addrm functions. * - * called with parent inode's i_mutex held + * LOCKING: + * Kernel thread context (may sleep). sysfs_mutex is locked on + * return. i_mutex of parent inode is locked on return if + * available. */ -int sysfs_dirent_exist(struct sysfs_dirent *parent_sd, - const unsigned char *new) +void sysfs_addrm_start(struct sysfs_addrm_cxt *acxt, + struct sysfs_dirent *parent_sd) { - struct sysfs_dirent * sd; + struct inode *inode; - list_for_each_entry(sd, &parent_sd->s_children, s_sibling) { - if (sd->s_element) { - const unsigned char *existing = sysfs_get_name(sd); - if (strcmp(existing, new)) - continue; - else - return -EEXIST; - } - } + memset(acxt, 0, sizeof(*acxt)); + acxt->parent_sd = parent_sd; - return 0; + /* Lookup parent inode. inode initialization and I_NEW + * clearing are protected by sysfs_mutex. By grabbing it and + * looking up with _nowait variant, inode state can be + * determined reliably. + */ + mutex_lock(&sysfs_mutex); + + inode = ilookup5_nowait(sysfs_sb, parent_sd->s_ino, sysfs_ilookup_test, + parent_sd); + + if (inode && !(inode->i_state & I_NEW)) { + /* parent inode available */ + acxt->parent_inode = inode; + + /* sysfs_mutex is below i_mutex in lock hierarchy. + * First, trylock i_mutex. If fails, unlock + * sysfs_mutex and lock them in order. + */ + if (!mutex_trylock(&inode->i_mutex)) { + mutex_unlock(&sysfs_mutex); + mutex_lock(&inode->i_mutex); + mutex_lock(&sysfs_mutex); + } + } else + iput(inode); } +/** + * sysfs_add_one - add sysfs_dirent to parent + * @acxt: addrm context to use + * @sd: sysfs_dirent to be added + * + * Get @acxt->parent_sd and set sd->s_parent to it and increment + * nlink of parent inode if @sd is a directory. @sd is NOT + * linked into the children list of the parent. The caller + * should invoke sysfs_link_sibling() after this function + * completes if @sd needs to be on the children list. + * + * This function should be called between calls to + * sysfs_addrm_start() and sysfs_addrm_finish() and should be + * passed the same @acxt as passed to sysfs_addrm_start(). + * + * LOCKING: + * Determined by sysfs_addrm_start(). + */ +void sysfs_add_one(struct sysfs_addrm_cxt *acxt, struct sysfs_dirent *sd) +{ + sd->s_parent = sysfs_get(acxt->parent_sd); + + if (sysfs_type(sd) == SYSFS_DIR && acxt->parent_inode) + inc_nlink(acxt->parent_inode); + + acxt->cnt++; +} -static struct sysfs_dirent * -__sysfs_make_dirent(struct dentry *dentry, void *element, mode_t mode, int type) +/** + * sysfs_remove_one - remove sysfs_dirent from parent + * @acxt: addrm context to use + * @sd: sysfs_dirent to be added + * + * Mark @sd removed and drop nlink of parent inode if @sd is a + * directory. @sd is NOT unlinked from the children list of the + * parent. The caller is repsonsible for removing @sd from the + * children list before calling this function. + * + * This function should be called between calls to + * sysfs_addrm_start() and sysfs_addrm_finish() and should be + * passed the same @acxt as passed to sysfs_addrm_start(). + * + * LOCKING: + * Determined by sysfs_addrm_start(). + */ +void sysfs_remove_one(struct sysfs_addrm_cxt *acxt, struct sysfs_dirent *sd) { - struct sysfs_dirent * sd; + BUG_ON(sd->s_sibling || (sd->s_flags & SYSFS_FLAG_REMOVED)); - sd = __sysfs_new_dirent(element); - if (!sd) - goto out; + sd->s_flags |= SYSFS_FLAG_REMOVED; + sd->s_sibling = acxt->removed; + acxt->removed = sd; - sd->s_mode = mode; - sd->s_type = type; - sd->s_dentry = dentry; - if (dentry) { - dentry->d_fsdata = sysfs_get(sd); - dentry->d_op = &sysfs_dentry_ops; - } + if (sysfs_type(sd) == SYSFS_DIR && acxt->parent_inode) + drop_nlink(acxt->parent_inode); -out: - return sd; + acxt->cnt++; } -int sysfs_make_dirent(struct sysfs_dirent * parent_sd, struct dentry * dentry, - void * element, umode_t mode, int type) +/** + * sysfs_drop_dentry - drop dentry for the specified sysfs_dirent + * @sd: target sysfs_dirent + * + * Drop dentry for @sd. @sd must have been unlinked from its + * parent on entry to this function such that it can't be looked + * up anymore. + * + * @sd->s_dentry which is protected with sysfs_assoc_lock points + * to the currently associated dentry but we're not holding a + * reference to it and racing with dput(). Grab dcache_lock and + * verify dentry before dropping it. If @sd->s_dentry is NULL or + * dput() beats us, no need to bother. + */ +static void sysfs_drop_dentry(struct sysfs_dirent *sd) { - struct sysfs_dirent *sd; + struct dentry *dentry = NULL; + struct inode *inode; + + /* We're not holding a reference to ->s_dentry dentry but the + * field will stay valid as long as sysfs_assoc_lock is held. + */ + spin_lock(&sysfs_assoc_lock); + spin_lock(&dcache_lock); + + /* drop dentry if it's there and dput() didn't kill it yet */ + if (sd->s_dentry && sd->s_dentry->d_inode) { + dentry = dget_locked(sd->s_dentry); + spin_lock(&dentry->d_lock); + __d_drop(dentry); + spin_unlock(&dentry->d_lock); + } - sd = __sysfs_make_dirent(dentry, element, mode, type); - __sysfs_list_dirent(parent_sd, sd); + spin_unlock(&dcache_lock); + spin_unlock(&sysfs_assoc_lock); - return sd ? 0 : -ENOMEM; + /* dentries for shadowed inodes are pinned, unpin */ + if (dentry && sysfs_is_shadowed_inode(dentry->d_inode)) + dput(dentry); + dput(dentry); + + /* adjust nlink and update timestamp */ + inode = ilookup(sysfs_sb, sd->s_ino); + if (inode) { + mutex_lock(&inode->i_mutex); + + inode->i_ctime = CURRENT_TIME; + drop_nlink(inode); + if (sysfs_type(sd) == SYSFS_DIR) + drop_nlink(inode); + + mutex_unlock(&inode->i_mutex); + iput(inode); + } } -static int init_dir(struct inode * inode) +/** + * sysfs_addrm_finish - finish up sysfs_dirent add/remove + * @acxt: addrm context to finish up + * + * Finish up sysfs_dirent add/remove. Resources acquired by + * sysfs_addrm_start() are released and removed sysfs_dirents are + * cleaned up. Timestamps on the parent inode are updated. + * + * LOCKING: + * All mutexes acquired by sysfs_addrm_start() are released. + * + * RETURNS: + * Number of added/removed sysfs_dirents since sysfs_addrm_start(). + */ +int sysfs_addrm_finish(struct sysfs_addrm_cxt *acxt) { - inode->i_op = &sysfs_dir_inode_operations; - inode->i_fop = &sysfs_dir_operations; + /* release resources acquired by sysfs_addrm_start() */ + mutex_unlock(&sysfs_mutex); + if (acxt->parent_inode) { + struct inode *inode = acxt->parent_inode; - /* directory inodes start off with i_nlink == 2 (for "." entry) */ - inc_nlink(inode); - return 0; + /* if added/removed, update timestamps on the parent */ + if (acxt->cnt) + inode->i_ctime = inode->i_mtime = CURRENT_TIME; + + mutex_unlock(&inode->i_mutex); + iput(inode); + } + + /* kill removed sysfs_dirents */ + while (acxt->removed) { + struct sysfs_dirent *sd = acxt->removed; + + acxt->removed = sd->s_sibling; + sd->s_sibling = NULL; + + sysfs_drop_dentry(sd); + sysfs_deactivate(sd); + sysfs_put(sd); + } + + return acxt->cnt; } -static int init_file(struct inode * inode) +/** + * sysfs_find_dirent - find sysfs_dirent with the given name + * @parent_sd: sysfs_dirent to search under + * @name: name to look for + * + * Look for sysfs_dirent with name @name under @parent_sd. + * + * LOCKING: + * mutex_lock(sysfs_mutex) + * + * RETURNS: + * Pointer to sysfs_dirent if found, NULL if not. + */ +struct sysfs_dirent *sysfs_find_dirent(struct sysfs_dirent *parent_sd, + const unsigned char *name) { - inode->i_size = PAGE_SIZE; - inode->i_fop = &sysfs_file_operations; - return 0; + struct sysfs_dirent *sd; + + for (sd = parent_sd->s_children; sd; sd = sd->s_sibling) + if (sysfs_type(sd) && !strcmp(sd->s_name, name)) + return sd; + return NULL; } -static int init_symlink(struct inode * inode) +/** + * sysfs_get_dirent - find and get sysfs_dirent with the given name + * @parent_sd: sysfs_dirent to search under + * @name: name to look for + * + * Look for sysfs_dirent with name @name under @parent_sd and get + * it if found. + * + * LOCKING: + * Kernel thread context (may sleep). Grabs sysfs_mutex. + * + * RETURNS: + * Pointer to sysfs_dirent if found, NULL if not. + */ +struct sysfs_dirent *sysfs_get_dirent(struct sysfs_dirent *parent_sd, + const unsigned char *name) { - inode->i_op = &sysfs_symlink_inode_operations; - return 0; + struct sysfs_dirent *sd; + + mutex_lock(&sysfs_mutex); + sd = sysfs_find_dirent(parent_sd, name); + sysfs_get(sd); + mutex_unlock(&sysfs_mutex); + + return sd; } -static int create_dir(struct kobject * k, struct dentry * p, - const char * n, struct dentry ** d) +static int create_dir(struct kobject *kobj, struct sysfs_dirent *parent_sd, + const char *name, struct sysfs_dirent **p_sd) { - int error; umode_t mode = S_IFDIR| S_IRWXU | S_IRUGO | S_IXUGO; + struct sysfs_addrm_cxt acxt; + struct sysfs_dirent *sd; - mutex_lock(&p->d_inode->i_mutex); - *d = lookup_one_len(n, p, strlen(n)); - if (!IS_ERR(*d)) { - if (sysfs_dirent_exist(p->d_fsdata, n)) - error = -EEXIST; - else - error = sysfs_make_dirent(p->d_fsdata, *d, k, mode, - SYSFS_DIR); - if (!error) { - error = sysfs_create(*d, mode, init_dir); - if (!error) { - inc_nlink(p->d_inode); - (*d)->d_op = &sysfs_dentry_ops; - d_rehash(*d); - } - } - if (error && (error != -EEXIST)) { - struct sysfs_dirent *sd = (*d)->d_fsdata; - if (sd) { - list_del_init(&sd->s_sibling); - sysfs_put(sd); - } - d_drop(*d); - } - dput(*d); - } else - error = PTR_ERR(*d); - mutex_unlock(&p->d_inode->i_mutex); - return error; -} + /* allocate */ + sd = sysfs_new_dirent(name, mode, SYSFS_DIR); + if (!sd) + return -ENOMEM; + sd->s_elem.dir.kobj = kobj; + /* link in */ + sysfs_addrm_start(&acxt, parent_sd); + if (!sysfs_find_dirent(parent_sd, name)) { + sysfs_add_one(&acxt, sd); + sysfs_link_sibling(sd); + } + if (sysfs_addrm_finish(&acxt)) { + *p_sd = sd; + return 0; + } -int sysfs_create_subdir(struct kobject * k, const char * n, struct dentry ** d) + sysfs_put(sd); + return -EEXIST; +} + +int sysfs_create_subdir(struct kobject *kobj, const char *name, + struct sysfs_dirent **p_sd) { - return create_dir(k,k->dentry,n,d); + return create_dir(kobj, kobj->sd, name, p_sd); } /** * sysfs_create_dir - create a directory for an object. * @kobj: object we're creating directory for. - * @shadow_parent: parent parent object. + * @shadow_parent: parent object. */ - -int sysfs_create_dir(struct kobject * kobj, struct dentry *shadow_parent) +int sysfs_create_dir(struct kobject *kobj, + struct sysfs_dirent *shadow_parent_sd) { - struct dentry * dentry = NULL; - struct dentry * parent; + struct sysfs_dirent *parent_sd, *sd; int error = 0; BUG_ON(!kobj); - if (shadow_parent) - parent = shadow_parent; + if (shadow_parent_sd) + parent_sd = shadow_parent_sd; else if (kobj->parent) - parent = kobj->parent->dentry; + parent_sd = kobj->parent->sd; else if (sysfs_mount && sysfs_mount->mnt_sb) - parent = sysfs_mount->mnt_sb->s_root; + parent_sd = sysfs_mount->mnt_sb->s_root->d_fsdata; else return -EFAULT; - error = create_dir(kobj,parent,kobject_name(kobj),&dentry); + error = create_dir(kobj, parent_sd, kobject_name(kobj), &sd); if (!error) - kobj->dentry = dentry; + kobj->sd = sd; return error; } -/* attaches attribute's sysfs_dirent to the dentry corresponding to the - * attribute file - */ -static int sysfs_attach_attr(struct sysfs_dirent * sd, struct dentry * dentry) +static int sysfs_count_nlink(struct sysfs_dirent *sd) { - struct attribute * attr = NULL; - struct bin_attribute * bin_attr = NULL; - int (* init) (struct inode *) = NULL; - int error = 0; - - if (sd->s_type & SYSFS_KOBJ_BIN_ATTR) { - bin_attr = sd->s_element; - attr = &bin_attr->attr; - } else { - attr = sd->s_element; - init = init_file; - } + struct sysfs_dirent *child; + int nr = 0; - dentry->d_fsdata = sysfs_get(sd); - /* protect sd->s_dentry against sysfs_d_iput */ - spin_lock(&sysfs_lock); - sd->s_dentry = dentry; - spin_unlock(&sysfs_lock); - error = sysfs_create(dentry, (attr->mode & S_IALLUGO) | S_IFREG, init); - if (error) { - sysfs_put(sd); - return error; - } - - if (bin_attr) { - dentry->d_inode->i_size = bin_attr->size; - dentry->d_inode->i_fop = &bin_fops; - } - dentry->d_op = &sysfs_dentry_ops; - d_rehash(dentry); - - return 0; -} - -static int sysfs_attach_link(struct sysfs_dirent * sd, struct dentry * dentry) -{ - int err = 0; - - dentry->d_fsdata = sysfs_get(sd); - /* protect sd->s_dentry against sysfs_d_iput */ - spin_lock(&sysfs_lock); - sd->s_dentry = dentry; - spin_unlock(&sysfs_lock); - err = sysfs_create(dentry, S_IFLNK|S_IRWXUGO, init_symlink); - if (!err) { - dentry->d_op = &sysfs_dentry_ops; - d_rehash(dentry); - } else - sysfs_put(sd); - - return err; + for (child = sd->s_children; child; child = child->s_sibling) + if (sysfs_type(child) == SYSFS_DIR) + nr++; + return nr + 2; } static struct dentry * sysfs_lookup(struct inode *dir, struct dentry *dentry, @@ -303,24 +761,60 @@ static struct dentry * sysfs_lookup(struct inode *dir, struct dentry *dentry, { struct sysfs_dirent * parent_sd = dentry->d_parent->d_fsdata; struct sysfs_dirent * sd; - int err = 0; + struct bin_attribute *bin_attr; + struct inode *inode; + int found = 0; - list_for_each_entry(sd, &parent_sd->s_children, s_sibling) { - if (sd->s_type & SYSFS_NOT_PINNED) { - const unsigned char * name = sysfs_get_name(sd); + for (sd = parent_sd->s_children; sd; sd = sd->s_sibling) { + if (sysfs_type(sd) && + !strcmp(sd->s_name, dentry->d_name.name)) { + found = 1; + break; + } + } - if (strcmp(name, dentry->d_name.name)) - continue; + /* no such entry */ + if (!found) + return NULL; - if (sd->s_type & SYSFS_KOBJ_LINK) - err = sysfs_attach_link(sd, dentry); - else - err = sysfs_attach_attr(sd, dentry); + /* attach dentry and inode */ + inode = sysfs_get_inode(sd); + if (!inode) + return ERR_PTR(-ENOMEM); + + mutex_lock(&sysfs_mutex); + + if (inode->i_state & I_NEW) { + /* initialize inode according to type */ + switch (sysfs_type(sd)) { + case SYSFS_DIR: + inode->i_op = &sysfs_dir_inode_operations; + inode->i_fop = &sysfs_dir_operations; + inode->i_nlink = sysfs_count_nlink(sd); + break; + case SYSFS_KOBJ_ATTR: + inode->i_size = PAGE_SIZE; + inode->i_fop = &sysfs_file_operations; + break; + case SYSFS_KOBJ_BIN_ATTR: + bin_attr = sd->s_elem.bin_attr.bin_attr; + inode->i_size = bin_attr->size; + inode->i_fop = &bin_fops; break; + case SYSFS_KOBJ_LINK: + inode->i_op = &sysfs_symlink_inode_operations; + break; + default: + BUG(); } } - return ERR_PTR(err); + sysfs_instantiate(dentry, inode); + sysfs_attach_dentry(sd, dentry); + + mutex_unlock(&sysfs_mutex); + + return NULL; } const struct inode_operations sysfs_dir_inode_operations = { @@ -328,58 +822,46 @@ const struct inode_operations sysfs_dir_inode_operations = { .setattr = sysfs_setattr, }; -static void remove_dir(struct dentry * d) +static void remove_dir(struct sysfs_dirent *sd) { - struct dentry * parent = dget(d->d_parent); - struct sysfs_dirent * sd; - - mutex_lock(&parent->d_inode->i_mutex); - d_delete(d); - sd = d->d_fsdata; - list_del_init(&sd->s_sibling); - sysfs_put(sd); - if (d->d_inode) - simple_rmdir(parent->d_inode,d); - - pr_debug(" o %s removing done (%d)\n",d->d_name.name, - atomic_read(&d->d_count)); + struct sysfs_addrm_cxt acxt; - mutex_unlock(&parent->d_inode->i_mutex); - dput(parent); + sysfs_addrm_start(&acxt, sd->s_parent); + sysfs_unlink_sibling(sd); + sysfs_remove_one(&acxt, sd); + sysfs_addrm_finish(&acxt); } -void sysfs_remove_subdir(struct dentry * d) +void sysfs_remove_subdir(struct sysfs_dirent *sd) { - remove_dir(d); + remove_dir(sd); } -static void __sysfs_remove_dir(struct dentry *dentry) +static void __sysfs_remove_dir(struct sysfs_dirent *dir_sd) { - struct sysfs_dirent * parent_sd; - struct sysfs_dirent * sd, * tmp; + struct sysfs_addrm_cxt acxt; + struct sysfs_dirent **pos; - dget(dentry); - if (!dentry) + if (!dir_sd) return; - pr_debug("sysfs %s: removing dir\n",dentry->d_name.name); - mutex_lock(&dentry->d_inode->i_mutex); - parent_sd = dentry->d_fsdata; - list_for_each_entry_safe(sd, tmp, &parent_sd->s_children, s_sibling) { - if (!sd->s_element || !(sd->s_type & SYSFS_NOT_PINNED)) - continue; - list_del_init(&sd->s_sibling); - sysfs_drop_dentry(sd, dentry); - sysfs_put(sd); + pr_debug("sysfs %s: removing dir\n", dir_sd->s_name); + sysfs_addrm_start(&acxt, dir_sd); + pos = &dir_sd->s_children; + while (*pos) { + struct sysfs_dirent *sd = *pos; + + if (sysfs_type(sd) && sysfs_type(sd) != SYSFS_DIR) { + *pos = sd->s_sibling; + sd->s_sibling = NULL; + sysfs_remove_one(&acxt, sd); + } else + pos = &(*pos)->s_sibling; } - mutex_unlock(&dentry->d_inode->i_mutex); + sysfs_addrm_finish(&acxt); - remove_dir(dentry); - /** - * Drop reference from dget() on entrance. - */ - dput(dentry); + remove_dir(dir_sd); } /** @@ -393,102 +875,166 @@ static void __sysfs_remove_dir(struct dentry *dentry) void sysfs_remove_dir(struct kobject * kobj) { - __sysfs_remove_dir(kobj->dentry); - kobj->dentry = NULL; + struct sysfs_dirent *sd = kobj->sd; + + spin_lock(&sysfs_assoc_lock); + kobj->sd = NULL; + spin_unlock(&sysfs_assoc_lock); + + __sysfs_remove_dir(sd); } -int sysfs_rename_dir(struct kobject * kobj, struct dentry *new_parent, +int sysfs_rename_dir(struct kobject *kobj, struct sysfs_dirent *new_parent_sd, const char *new_name) { - int error = 0; - struct dentry * new_dentry; + struct sysfs_dirent *sd = kobj->sd; + struct dentry *new_parent = NULL; + struct dentry *old_dentry = NULL, *new_dentry = NULL; + const char *dup_name = NULL; + int error; - if (!new_parent) - return -EFAULT; + /* get dentries */ + old_dentry = sysfs_get_dentry(sd); + if (IS_ERR(old_dentry)) { + error = PTR_ERR(old_dentry); + goto out_dput; + } - down_write(&sysfs_rename_sem); + new_parent = sysfs_get_dentry(new_parent_sd); + if (IS_ERR(new_parent)) { + error = PTR_ERR(new_parent); + goto out_dput; + } + + /* lock new_parent and get dentry for new name */ mutex_lock(&new_parent->d_inode->i_mutex); new_dentry = lookup_one_len(new_name, new_parent, strlen(new_name)); - if (!IS_ERR(new_dentry)) { - /* By allowing two different directories with the - * same d_parent we allow this routine to move - * between different shadows of the same directory - */ - if (kobj->dentry->d_parent->d_inode != new_parent->d_inode) - return -EINVAL; - else if (new_dentry->d_parent->d_inode != new_parent->d_inode) - error = -EINVAL; - else if (new_dentry == kobj->dentry) - error = -EINVAL; - else if (!new_dentry->d_inode) { - error = kobject_set_name(kobj, "%s", new_name); - if (!error) { - struct sysfs_dirent *sd, *parent_sd; - - d_add(new_dentry, NULL); - d_move(kobj->dentry, new_dentry); - - sd = kobj->dentry->d_fsdata; - parent_sd = new_parent->d_fsdata; - - list_del_init(&sd->s_sibling); - list_add(&sd->s_sibling, &parent_sd->s_children); - } - else - d_drop(new_dentry); - } else - error = -EEXIST; - dput(new_dentry); + if (IS_ERR(new_dentry)) { + error = PTR_ERR(new_dentry); + goto out_unlock; } - mutex_unlock(&new_parent->d_inode->i_mutex); - up_write(&sysfs_rename_sem); + /* By allowing two different directories with the same + * d_parent we allow this routine to move between different + * shadows of the same directory + */ + error = -EINVAL; + if (old_dentry->d_parent->d_inode != new_parent->d_inode || + new_dentry->d_parent->d_inode != new_parent->d_inode || + old_dentry == new_dentry) + goto out_unlock; + + error = -EEXIST; + if (new_dentry->d_inode) + goto out_unlock; + + /* rename kobject and sysfs_dirent */ + error = -ENOMEM; + new_name = dup_name = kstrdup(new_name, GFP_KERNEL); + if (!new_name) + goto out_drop; + + error = kobject_set_name(kobj, "%s", new_name); + if (error) + goto out_drop; + + dup_name = sd->s_name; + sd->s_name = new_name; + + /* move under the new parent */ + d_add(new_dentry, NULL); + d_move(sd->s_dentry, new_dentry); + + mutex_lock(&sysfs_mutex); + + sysfs_unlink_sibling(sd); + sysfs_get(new_parent_sd); + sysfs_put(sd->s_parent); + sd->s_parent = new_parent_sd; + sysfs_link_sibling(sd); + + mutex_unlock(&sysfs_mutex); + + error = 0; + goto out_unlock; + + out_drop: + d_drop(new_dentry); + out_unlock: + mutex_unlock(&new_parent->d_inode->i_mutex); + out_dput: + kfree(dup_name); + dput(new_parent); + dput(old_dentry); + dput(new_dentry); return error; } -int sysfs_move_dir(struct kobject *kobj, struct kobject *new_parent) +int sysfs_move_dir(struct kobject *kobj, struct kobject *new_parent_kobj) { - struct dentry *old_parent_dentry, *new_parent_dentry, *new_dentry; - struct sysfs_dirent *new_parent_sd, *sd; + struct sysfs_dirent *sd = kobj->sd; + struct sysfs_dirent *new_parent_sd; + struct dentry *old_parent, *new_parent = NULL; + struct dentry *old_dentry = NULL, *new_dentry = NULL; int error; - old_parent_dentry = kobj->parent ? - kobj->parent->dentry : sysfs_mount->mnt_sb->s_root; - new_parent_dentry = new_parent ? - new_parent->dentry : sysfs_mount->mnt_sb->s_root; + BUG_ON(!sd->s_parent); + new_parent_sd = new_parent_kobj->sd ? new_parent_kobj->sd : &sysfs_root; + + /* get dentries */ + old_dentry = sysfs_get_dentry(sd); + if (IS_ERR(old_dentry)) { + error = PTR_ERR(old_dentry); + goto out_dput; + } + old_parent = sd->s_parent->s_dentry; + + new_parent = sysfs_get_dentry(new_parent_sd); + if (IS_ERR(new_parent)) { + error = PTR_ERR(new_parent); + goto out_dput; + } - if (old_parent_dentry->d_inode == new_parent_dentry->d_inode) - return 0; /* nothing to move */ + if (old_parent->d_inode == new_parent->d_inode) { + error = 0; + goto out_dput; /* nothing to move */ + } again: - mutex_lock(&old_parent_dentry->d_inode->i_mutex); - if (!mutex_trylock(&new_parent_dentry->d_inode->i_mutex)) { - mutex_unlock(&old_parent_dentry->d_inode->i_mutex); + mutex_lock(&old_parent->d_inode->i_mutex); + if (!mutex_trylock(&new_parent->d_inode->i_mutex)) { + mutex_unlock(&old_parent->d_inode->i_mutex); goto again; } - new_parent_sd = new_parent_dentry->d_fsdata; - sd = kobj->dentry->d_fsdata; - - new_dentry = lookup_one_len(kobj->name, new_parent_dentry, - strlen(kobj->name)); + new_dentry = lookup_one_len(kobj->name, new_parent, strlen(kobj->name)); if (IS_ERR(new_dentry)) { error = PTR_ERR(new_dentry); - goto out; + goto out_unlock; } else error = 0; d_add(new_dentry, NULL); - d_move(kobj->dentry, new_dentry); + d_move(sd->s_dentry, new_dentry); dput(new_dentry); /* Remove from old parent's list and insert into new parent's list. */ - list_del_init(&sd->s_sibling); - list_add(&sd->s_sibling, &new_parent_sd->s_children); + mutex_lock(&sysfs_mutex); + + sysfs_unlink_sibling(sd); + sysfs_get(new_parent_sd); + sysfs_put(sd->s_parent); + sd->s_parent = new_parent_sd; + sysfs_link_sibling(sd); -out: - mutex_unlock(&new_parent_dentry->d_inode->i_mutex); - mutex_unlock(&old_parent_dentry->d_inode->i_mutex); + mutex_unlock(&sysfs_mutex); + out_unlock: + mutex_unlock(&new_parent->d_inode->i_mutex); + mutex_unlock(&old_parent->d_inode->i_mutex); + out_dput: + dput(new_parent); + dput(old_dentry); + dput(new_dentry); return error; } @@ -496,23 +1042,27 @@ static int sysfs_dir_open(struct inode *inode, struct file *file) { struct dentry * dentry = file->f_path.dentry; struct sysfs_dirent * parent_sd = dentry->d_fsdata; + struct sysfs_dirent * sd; - mutex_lock(&dentry->d_inode->i_mutex); - file->private_data = sysfs_new_dirent(parent_sd, NULL); - mutex_unlock(&dentry->d_inode->i_mutex); - - return file->private_data ? 0 : -ENOMEM; + sd = sysfs_new_dirent("_DIR_", 0, 0); + if (sd) { + mutex_lock(&sysfs_mutex); + sd->s_parent = sysfs_get(parent_sd); + sysfs_link_sibling(sd); + mutex_unlock(&sysfs_mutex); + } + file->private_data = sd; + return sd ? 0 : -ENOMEM; } static int sysfs_dir_close(struct inode *inode, struct file *file) { - struct dentry * dentry = file->f_path.dentry; struct sysfs_dirent * cursor = file->private_data; - mutex_lock(&dentry->d_inode->i_mutex); - list_del_init(&cursor->s_sibling); - mutex_unlock(&dentry->d_inode->i_mutex); + mutex_lock(&sysfs_mutex); + sysfs_unlink_sibling(cursor); + mutex_unlock(&sysfs_mutex); release_sysfs_dirent(cursor); @@ -530,7 +1080,7 @@ static int sysfs_readdir(struct file * filp, void * dirent, filldir_t filldir) struct dentry *dentry = filp->f_path.dentry; struct sysfs_dirent * parent_sd = dentry->d_fsdata; struct sysfs_dirent *cursor = filp->private_data; - struct list_head *p, *q = &cursor->s_sibling; + struct sysfs_dirent **pos; ino_t ino; int i = filp->f_pos; @@ -543,38 +1093,52 @@ static int sysfs_readdir(struct file * filp, void * dirent, filldir_t filldir) i++; /* fallthrough */ case 1: - ino = parent_ino(dentry); + if (parent_sd->s_parent) + ino = parent_sd->s_parent->s_ino; + else + ino = parent_sd->s_ino; if (filldir(dirent, "..", 2, i, ino, DT_DIR) < 0) break; filp->f_pos++; i++; /* fallthrough */ default: + mutex_lock(&sysfs_mutex); + + pos = &parent_sd->s_children; + while (*pos != cursor) + pos = &(*pos)->s_sibling; + + /* unlink cursor */ + *pos = cursor->s_sibling; + if (filp->f_pos == 2) - list_move(q, &parent_sd->s_children); + pos = &parent_sd->s_children; - for (p=q->next; p!= &parent_sd->s_children; p=p->next) { - struct sysfs_dirent *next; + for ( ; *pos; pos = &(*pos)->s_sibling) { + struct sysfs_dirent *next = *pos; const char * name; int len; - next = list_entry(p, struct sysfs_dirent, - s_sibling); - if (!next->s_element) + if (!sysfs_type(next)) continue; - name = sysfs_get_name(next); + name = next->s_name; len = strlen(name); ino = next->s_ino; if (filldir(dirent, name, len, filp->f_pos, ino, dt_type(next)) < 0) - return 0; + break; - list_move(q, p); - p = q; filp->f_pos++; } + + /* put cursor back in */ + cursor->s_sibling = *pos; + *pos = cursor; + + mutex_unlock(&sysfs_mutex); } return 0; } @@ -583,7 +1147,6 @@ static loff_t sysfs_dir_lseek(struct file * file, loff_t offset, int origin) { struct dentry * dentry = file->f_path.dentry; - mutex_lock(&dentry->d_inode->i_mutex); switch (origin) { case 1: offset += file->f_pos; @@ -591,31 +1154,35 @@ static loff_t sysfs_dir_lseek(struct file * file, loff_t offset, int origin) if (offset >= 0) break; default: - mutex_unlock(&file->f_path.dentry->d_inode->i_mutex); return -EINVAL; } if (offset != file->f_pos) { + mutex_lock(&sysfs_mutex); + file->f_pos = offset; if (file->f_pos >= 2) { struct sysfs_dirent *sd = dentry->d_fsdata; struct sysfs_dirent *cursor = file->private_data; - struct list_head *p; + struct sysfs_dirent **pos; loff_t n = file->f_pos - 2; - list_del(&cursor->s_sibling); - p = sd->s_children.next; - while (n && p != &sd->s_children) { - struct sysfs_dirent *next; - next = list_entry(p, struct sysfs_dirent, - s_sibling); - if (next->s_element) + sysfs_unlink_sibling(cursor); + + pos = &sd->s_children; + while (n && *pos) { + struct sysfs_dirent *next = *pos; + if (sysfs_type(next)) n--; - p = p->next; + pos = &(*pos)->s_sibling; } - list_add_tail(&cursor->s_sibling, p); + + cursor->s_sibling = *pos; + *pos = cursor; } + + mutex_unlock(&sysfs_mutex); } - mutex_unlock(&dentry->d_inode->i_mutex); + return offset; } @@ -628,12 +1195,20 @@ static loff_t sysfs_dir_lseek(struct file * file, loff_t offset, int origin) int sysfs_make_shadowed_dir(struct kobject *kobj, void * (*follow_link)(struct dentry *, struct nameidata *)) { + struct dentry *dentry; struct inode *inode; struct inode_operations *i_op; - inode = kobj->dentry->d_inode; - if (inode->i_op != &sysfs_dir_inode_operations) + /* get dentry for @kobj->sd, dentry of a shadowed dir is pinned */ + dentry = sysfs_get_dentry(kobj->sd); + if (IS_ERR(dentry)) + return PTR_ERR(dentry); + + inode = dentry->d_inode; + if (inode->i_op != &sysfs_dir_inode_operations) { + dput(dentry); return -EINVAL; + } i_op = kmalloc(sizeof(*i_op), GFP_KERNEL); if (!i_op) @@ -658,54 +1233,72 @@ int sysfs_make_shadowed_dir(struct kobject *kobj, * directory. */ -struct dentry *sysfs_create_shadow_dir(struct kobject *kobj) +struct sysfs_dirent *sysfs_create_shadow_dir(struct kobject *kobj) { - struct sysfs_dirent *sd; - struct dentry *parent, *dir, *shadow; + struct sysfs_dirent *parent_sd = kobj->sd->s_parent; + struct dentry *dir, *parent, *shadow; struct inode *inode; + struct sysfs_dirent *sd; + struct sysfs_addrm_cxt acxt; - dir = kobj->dentry; - inode = dir->d_inode; + dir = sysfs_get_dentry(kobj->sd); + if (IS_ERR(dir)) { + sd = (void *)dir; + goto out; + } parent = dir->d_parent; - shadow = ERR_PTR(-EINVAL); + + inode = dir->d_inode; + sd = ERR_PTR(-EINVAL); if (!sysfs_is_shadowed_inode(inode)) - goto out; + goto out_dput; shadow = d_alloc(parent, &dir->d_name); if (!shadow) goto nomem; - sd = __sysfs_make_dirent(shadow, kobj, inode->i_mode, SYSFS_DIR); + sd = sysfs_new_dirent("_SHADOW_", inode->i_mode, SYSFS_DIR); if (!sd) goto nomem; + sd->s_elem.dir.kobj = kobj; + sysfs_addrm_start(&acxt, parent_sd); + + /* add but don't link into children list */ + sysfs_add_one(&acxt, sd); + + /* attach and instantiate dentry */ + sysfs_attach_dentry(sd, shadow); d_instantiate(shadow, igrab(inode)); - inc_nlink(inode); - inc_nlink(parent->d_inode); - shadow->d_op = &sysfs_dentry_ops; + inc_nlink(inode); /* tj: synchronization? */ + + sysfs_addrm_finish(&acxt); dget(shadow); /* Extra count - pin the dentry in core */ -out: - return shadow; -nomem: + goto out_dput; + + nomem: dput(shadow); - shadow = ERR_PTR(-ENOMEM); - goto out; + sd = ERR_PTR(-ENOMEM); + out_dput: + dput(dir); + out: + return sd; } /** * sysfs_remove_shadow_dir - remove an object's directory. - * @shadow: dentry of shadow directory + * @shadow_sd: sysfs_dirent of shadow directory * * The only thing special about this is that we remove any files in * the directory before we remove the directory, and we've inlined * what used to be sysfs_rmdir() below, instead of calling separately. */ -void sysfs_remove_shadow_dir(struct dentry *shadow) +void sysfs_remove_shadow_dir(struct sysfs_dirent *shadow_sd) { - __sysfs_remove_dir(shadow); + __sysfs_remove_dir(shadow_sd); } const struct file_operations sysfs_dir_operations = { diff --git a/fs/sysfs/file.c b/fs/sysfs/file.c index b502c7197ec0..cc497994b2a8 100644 --- a/fs/sysfs/file.c +++ b/fs/sysfs/file.c @@ -50,29 +50,15 @@ static struct sysfs_ops subsys_sysfs_ops = { .store = subsys_attr_store, }; -/** - * add_to_collection - add buffer to a collection - * @buffer: buffer to be added - * @node: inode of set to add to - */ - -static inline void -add_to_collection(struct sysfs_buffer *buffer, struct inode *node) -{ - struct sysfs_buffer_collection *set = node->i_private; - - mutex_lock(&node->i_mutex); - list_add(&buffer->associates, &set->associates); - mutex_unlock(&node->i_mutex); -} - -static inline void -remove_from_collection(struct sysfs_buffer *buffer, struct inode *node) -{ - mutex_lock(&node->i_mutex); - list_del(&buffer->associates); - mutex_unlock(&node->i_mutex); -} +struct sysfs_buffer { + size_t count; + loff_t pos; + char * page; + struct sysfs_ops * ops; + struct semaphore sem; + int needs_read_fill; + int event; +}; /** * fill_read_buffer - allocate and fill buffer from object. @@ -87,9 +73,8 @@ remove_from_collection(struct sysfs_buffer *buffer, struct inode *node) */ static int fill_read_buffer(struct dentry * dentry, struct sysfs_buffer * buffer) { - struct sysfs_dirent * sd = dentry->d_fsdata; - struct attribute * attr = to_attr(dentry); - struct kobject * kobj = to_kobj(dentry->d_parent); + struct sysfs_dirent *attr_sd = dentry->d_fsdata; + struct kobject *kobj = attr_sd->s_parent->s_elem.dir.kobj; struct sysfs_ops * ops = buffer->ops; int ret = 0; ssize_t count; @@ -99,8 +84,15 @@ static int fill_read_buffer(struct dentry * dentry, struct sysfs_buffer * buffer if (!buffer->page) return -ENOMEM; - buffer->event = atomic_read(&sd->s_event); - count = ops->show(kobj,attr,buffer->page); + /* need attr_sd for attr and ops, its parent for kobj */ + if (!sysfs_get_active_two(attr_sd)) + return -ENODEV; + + buffer->event = atomic_read(&attr_sd->s_event); + count = ops->show(kobj, attr_sd->s_elem.attr.attr, buffer->page); + + sysfs_put_active_two(attr_sd); + BUG_ON(count > (ssize_t)PAGE_SIZE); if (count >= 0) { buffer->needs_read_fill = 0; @@ -138,10 +130,7 @@ sysfs_read_file(struct file *file, char __user *buf, size_t count, loff_t *ppos) down(&buffer->sem); if (buffer->needs_read_fill) { - if (buffer->orphaned) - retval = -ENODEV; - else - retval = fill_read_buffer(file->f_path.dentry,buffer); + retval = fill_read_buffer(file->f_path.dentry,buffer); if (retval) goto out; } @@ -196,14 +185,23 @@ fill_write_buffer(struct sysfs_buffer * buffer, const char __user * buf, size_t * passing the buffer that we acquired in fill_write_buffer(). */ -static int +static int flush_write_buffer(struct dentry * dentry, struct sysfs_buffer * buffer, size_t count) { - struct attribute * attr = to_attr(dentry); - struct kobject * kobj = to_kobj(dentry->d_parent); + struct sysfs_dirent *attr_sd = dentry->d_fsdata; + struct kobject *kobj = attr_sd->s_parent->s_elem.dir.kobj; struct sysfs_ops * ops = buffer->ops; + int rc; + + /* need attr_sd for attr and ops, its parent for kobj */ + if (!sysfs_get_active_two(attr_sd)) + return -ENODEV; + + rc = ops->store(kobj, attr_sd->s_elem.attr.attr, buffer->page, count); - return ops->store(kobj,attr,buffer->page,count); + sysfs_put_active_two(attr_sd); + + return rc; } @@ -231,37 +229,26 @@ sysfs_write_file(struct file *file, const char __user *buf, size_t count, loff_t ssize_t len; down(&buffer->sem); - if (buffer->orphaned) { - len = -ENODEV; - goto out; - } len = fill_write_buffer(buffer, buf, count); if (len > 0) len = flush_write_buffer(file->f_path.dentry, buffer, len); if (len > 0) *ppos += len; -out: up(&buffer->sem); return len; } static int sysfs_open_file(struct inode *inode, struct file *file) { - struct kobject *kobj = sysfs_get_kobject(file->f_path.dentry->d_parent); - struct attribute * attr = to_attr(file->f_path.dentry); - struct sysfs_buffer_collection *set; + struct sysfs_dirent *attr_sd = file->f_path.dentry->d_fsdata; + struct kobject *kobj = attr_sd->s_parent->s_elem.dir.kobj; struct sysfs_buffer * buffer; struct sysfs_ops * ops = NULL; - int error = 0; - - if (!kobj || !attr) - goto Einval; + int error; - /* Grab the module reference for this attribute if we have one */ - if (!try_module_get(attr->owner)) { - error = -ENODEV; - goto Done; - } + /* need attr_sd for attr and ops, its parent for kobj */ + if (!sysfs_get_active_two(attr_sd)) + return -ENODEV; /* if the kobject has no ktype, then we assume that it is a subsystem * itself, and use ops for it. @@ -273,33 +260,21 @@ static int sysfs_open_file(struct inode *inode, struct file *file) else ops = &subsys_sysfs_ops; + error = -EACCES; + /* No sysfs operations, either from having no subsystem, * or the subsystem have no operations. */ if (!ops) - goto Eaccess; - - /* make sure we have a collection to add our buffers to */ - mutex_lock(&inode->i_mutex); - if (!(set = inode->i_private)) { - if (!(set = inode->i_private = kmalloc(sizeof(struct sysfs_buffer_collection), GFP_KERNEL))) { - error = -ENOMEM; - goto Done; - } else { - INIT_LIST_HEAD(&set->associates); - } - } - mutex_unlock(&inode->i_mutex); + goto err_out; /* File needs write support. * The inode's perms must say it's ok, * and we must have a store method. */ if (file->f_mode & FMODE_WRITE) { - if (!(inode->i_mode & S_IWUGO) || !ops->store) - goto Eaccess; - + goto err_out; } /* File needs read support. @@ -308,48 +283,38 @@ static int sysfs_open_file(struct inode *inode, struct file *file) */ if (file->f_mode & FMODE_READ) { if (!(inode->i_mode & S_IRUGO) || !ops->show) - goto Eaccess; + goto err_out; } /* No error? Great, allocate a buffer for the file, and store it * it in file->private_data for easy access. */ + error = -ENOMEM; buffer = kzalloc(sizeof(struct sysfs_buffer), GFP_KERNEL); - if (buffer) { - INIT_LIST_HEAD(&buffer->associates); - init_MUTEX(&buffer->sem); - buffer->needs_read_fill = 1; - buffer->ops = ops; - add_to_collection(buffer, inode); - file->private_data = buffer; - } else - error = -ENOMEM; - goto Done; - - Einval: - error = -EINVAL; - goto Done; - Eaccess: - error = -EACCES; - module_put(attr->owner); - Done: - if (error) - kobject_put(kobj); + if (!buffer) + goto err_out; + + init_MUTEX(&buffer->sem); + buffer->needs_read_fill = 1; + buffer->ops = ops; + file->private_data = buffer; + + /* open succeeded, put active references and pin attr_sd */ + sysfs_put_active_two(attr_sd); + sysfs_get(attr_sd); + return 0; + + err_out: + sysfs_put_active_two(attr_sd); return error; } static int sysfs_release(struct inode * inode, struct file * filp) { - struct kobject * kobj = to_kobj(filp->f_path.dentry->d_parent); - struct attribute * attr = to_attr(filp->f_path.dentry); - struct module * owner = attr->owner; - struct sysfs_buffer * buffer = filp->private_data; + struct sysfs_dirent *attr_sd = filp->f_path.dentry->d_fsdata; + struct sysfs_buffer *buffer = filp->private_data; - if (buffer) - remove_from_collection(buffer, inode); - kobject_put(kobj); - /* After this point, attr should not be accessed. */ - module_put(owner); + sysfs_put(attr_sd); if (buffer) { if (buffer->page) @@ -376,57 +341,43 @@ static int sysfs_release(struct inode * inode, struct file * filp) static unsigned int sysfs_poll(struct file *filp, poll_table *wait) { struct sysfs_buffer * buffer = filp->private_data; - struct kobject * kobj = to_kobj(filp->f_path.dentry->d_parent); - struct sysfs_dirent * sd = filp->f_path.dentry->d_fsdata; - int res = 0; + struct sysfs_dirent *attr_sd = filp->f_path.dentry->d_fsdata; + struct kobject *kobj = attr_sd->s_parent->s_elem.dir.kobj; + + /* need parent for the kobj, grab both */ + if (!sysfs_get_active_two(attr_sd)) + goto trigger; poll_wait(filp, &kobj->poll, wait); - if (buffer->event != atomic_read(&sd->s_event)) { - res = POLLERR|POLLPRI; - buffer->needs_read_fill = 1; - } + sysfs_put_active_two(attr_sd); - return res; -} + if (buffer->event != atomic_read(&attr_sd->s_event)) + goto trigger; + return 0; -static struct dentry *step_down(struct dentry *dir, const char * name) -{ - struct dentry * de; - - if (dir == NULL || dir->d_inode == NULL) - return NULL; - - mutex_lock(&dir->d_inode->i_mutex); - de = lookup_one_len(name, dir, strlen(name)); - mutex_unlock(&dir->d_inode->i_mutex); - dput(dir); - if (IS_ERR(de)) - return NULL; - if (de->d_inode == NULL) { - dput(de); - return NULL; - } - return de; + trigger: + buffer->needs_read_fill = 1; + return POLLERR|POLLPRI; } -void sysfs_notify(struct kobject * k, char *dir, char *attr) +void sysfs_notify(struct kobject *k, char *dir, char *attr) { - struct dentry *de = k->dentry; - if (de) - dget(de); - if (de && dir) - de = step_down(de, dir); - if (de && attr) - de = step_down(de, attr); - if (de) { - struct sysfs_dirent * sd = de->d_fsdata; - if (sd) - atomic_inc(&sd->s_event); + struct sysfs_dirent *sd = k->sd; + + mutex_lock(&sysfs_mutex); + + if (sd && dir) + sd = sysfs_find_dirent(sd, dir); + if (sd && attr) + sd = sysfs_find_dirent(sd, attr); + if (sd) { + atomic_inc(&sd->s_event); wake_up_interruptible(&k->poll); - dput(de); } + + mutex_unlock(&sysfs_mutex); } EXPORT_SYMBOL_GPL(sysfs_notify); @@ -440,19 +391,30 @@ const struct file_operations sysfs_file_operations = { }; -int sysfs_add_file(struct dentry * dir, const struct attribute * attr, int type) +int sysfs_add_file(struct sysfs_dirent *dir_sd, const struct attribute *attr, + int type) { - struct sysfs_dirent * parent_sd = dir->d_fsdata; umode_t mode = (attr->mode & S_IALLUGO) | S_IFREG; - int error = -EEXIST; + struct sysfs_addrm_cxt acxt; + struct sysfs_dirent *sd; - mutex_lock(&dir->d_inode->i_mutex); - if (!sysfs_dirent_exist(parent_sd, attr->name)) - error = sysfs_make_dirent(parent_sd, NULL, (void *)attr, - mode, type); - mutex_unlock(&dir->d_inode->i_mutex); + sd = sysfs_new_dirent(attr->name, mode, type); + if (!sd) + return -ENOMEM; + sd->s_elem.attr.attr = (void *)attr; - return error; + sysfs_addrm_start(&acxt, dir_sd); + + if (!sysfs_find_dirent(dir_sd, attr->name)) { + sysfs_add_one(&acxt, sd); + sysfs_link_sibling(sd); + } + + if (sysfs_addrm_finish(&acxt)) + return 0; + + sysfs_put(sd); + return -EEXIST; } @@ -464,9 +426,9 @@ int sysfs_add_file(struct dentry * dir, const struct attribute * attr, int type) int sysfs_create_file(struct kobject * kobj, const struct attribute * attr) { - BUG_ON(!kobj || !kobj->dentry || !attr); + BUG_ON(!kobj || !kobj->sd || !attr); - return sysfs_add_file(kobj->dentry, attr, SYSFS_KOBJ_ATTR); + return sysfs_add_file(kobj->sd, attr, SYSFS_KOBJ_ATTR); } @@ -480,16 +442,16 @@ int sysfs_create_file(struct kobject * kobj, const struct attribute * attr) int sysfs_add_file_to_group(struct kobject *kobj, const struct attribute *attr, const char *group) { - struct dentry *dir; + struct sysfs_dirent *dir_sd; int error; - dir = lookup_one_len(group, kobj->dentry, strlen(group)); - if (IS_ERR(dir)) - error = PTR_ERR(dir); - else { - error = sysfs_add_file(dir, attr, SYSFS_KOBJ_ATTR); - dput(dir); - } + dir_sd = sysfs_get_dirent(kobj->sd, group); + if (!dir_sd) + return -ENOENT; + + error = sysfs_add_file(dir_sd, attr, SYSFS_KOBJ_ATTR); + sysfs_put(dir_sd); + return error; } EXPORT_SYMBOL_GPL(sysfs_add_file_to_group); @@ -502,30 +464,31 @@ EXPORT_SYMBOL_GPL(sysfs_add_file_to_group); */ int sysfs_update_file(struct kobject * kobj, const struct attribute * attr) { - struct dentry * dir = kobj->dentry; - struct dentry * victim; - int res = -ENOENT; - - mutex_lock(&dir->d_inode->i_mutex); - victim = lookup_one_len(attr->name, dir, strlen(attr->name)); - if (!IS_ERR(victim)) { - /* make sure dentry is really there */ - if (victim->d_inode && - (victim->d_parent->d_inode == dir->d_inode)) { - victim->d_inode->i_mtime = CURRENT_TIME; - fsnotify_modify(victim); - res = 0; - } else - d_drop(victim); - - /** - * Drop the reference acquired from lookup_one_len() above. - */ - dput(victim); + struct sysfs_dirent *victim_sd = NULL; + struct dentry *victim = NULL; + int rc; + + rc = -ENOENT; + victim_sd = sysfs_get_dirent(kobj->sd, attr->name); + if (!victim_sd) + goto out; + + victim = sysfs_get_dentry(victim_sd); + if (IS_ERR(victim)) { + rc = PTR_ERR(victim); + victim = NULL; + goto out; } - mutex_unlock(&dir->d_inode->i_mutex); - return res; + mutex_lock(&victim->d_inode->i_mutex); + victim->d_inode->i_mtime = CURRENT_TIME; + fsnotify_modify(victim); + mutex_unlock(&victim->d_inode->i_mutex); + rc = 0; + out: + dput(victim); + sysfs_put(victim_sd); + return rc; } @@ -538,30 +501,34 @@ int sysfs_update_file(struct kobject * kobj, const struct attribute * attr) */ int sysfs_chmod_file(struct kobject *kobj, struct attribute *attr, mode_t mode) { - struct dentry *dir = kobj->dentry; - struct dentry *victim; + struct sysfs_dirent *victim_sd = NULL; + struct dentry *victim = NULL; struct inode * inode; struct iattr newattrs; - int res = -ENOENT; - - mutex_lock(&dir->d_inode->i_mutex); - victim = lookup_one_len(attr->name, dir, strlen(attr->name)); - if (!IS_ERR(victim)) { - if (victim->d_inode && - (victim->d_parent->d_inode == dir->d_inode)) { - inode = victim->d_inode; - mutex_lock(&inode->i_mutex); - newattrs.ia_mode = (mode & S_IALLUGO) | - (inode->i_mode & ~S_IALLUGO); - newattrs.ia_valid = ATTR_MODE | ATTR_CTIME; - res = notify_change(victim, &newattrs); - mutex_unlock(&inode->i_mutex); - } - dput(victim); + int rc; + + rc = -ENOENT; + victim_sd = sysfs_get_dirent(kobj->sd, attr->name); + if (!victim_sd) + goto out; + + victim = sysfs_get_dentry(victim_sd); + if (IS_ERR(victim)) { + rc = PTR_ERR(victim); + victim = NULL; + goto out; } - mutex_unlock(&dir->d_inode->i_mutex); - return res; + inode = victim->d_inode; + mutex_lock(&inode->i_mutex); + newattrs.ia_mode = (mode & S_IALLUGO) | (inode->i_mode & ~S_IALLUGO); + newattrs.ia_valid = ATTR_MODE | ATTR_CTIME; + rc = notify_change(victim, &newattrs); + mutex_unlock(&inode->i_mutex); + out: + dput(victim); + sysfs_put(victim_sd); + return rc; } EXPORT_SYMBOL_GPL(sysfs_chmod_file); @@ -576,7 +543,7 @@ EXPORT_SYMBOL_GPL(sysfs_chmod_file); void sysfs_remove_file(struct kobject * kobj, const struct attribute * attr) { - sysfs_hash_and_remove(kobj->dentry, attr->name); + sysfs_hash_and_remove(kobj->sd, attr->name); } @@ -589,12 +556,12 @@ void sysfs_remove_file(struct kobject * kobj, const struct attribute * attr) void sysfs_remove_file_from_group(struct kobject *kobj, const struct attribute *attr, const char *group) { - struct dentry *dir; + struct sysfs_dirent *dir_sd; - dir = lookup_one_len(group, kobj->dentry, strlen(group)); - if (!IS_ERR(dir)) { - sysfs_hash_and_remove(dir, attr->name); - dput(dir); + dir_sd = sysfs_get_dirent(kobj->sd, group); + if (dir_sd) { + sysfs_hash_and_remove(dir_sd, attr->name); + sysfs_put(dir_sd); } } EXPORT_SYMBOL_GPL(sysfs_remove_file_from_group); diff --git a/fs/sysfs/group.c b/fs/sysfs/group.c index 52eed2a7a5ef..f318b73c790c 100644 --- a/fs/sysfs/group.c +++ b/fs/sysfs/group.c @@ -18,26 +18,25 @@ #include "sysfs.h" -static void remove_files(struct dentry * dir, - const struct attribute_group * grp) +static void remove_files(struct sysfs_dirent *dir_sd, + const struct attribute_group *grp) { struct attribute *const* attr; for (attr = grp->attrs; *attr; attr++) - sysfs_hash_and_remove(dir,(*attr)->name); + sysfs_hash_and_remove(dir_sd, (*attr)->name); } -static int create_files(struct dentry * dir, - const struct attribute_group * grp) +static int create_files(struct sysfs_dirent *dir_sd, + const struct attribute_group *grp) { struct attribute *const* attr; int error = 0; - for (attr = grp->attrs; *attr && !error; attr++) { - error = sysfs_add_file(dir, *attr, SYSFS_KOBJ_ATTR); - } + for (attr = grp->attrs; *attr && !error; attr++) + error = sysfs_add_file(dir_sd, *attr, SYSFS_KOBJ_ATTR); if (error) - remove_files(dir,grp); + remove_files(dir_sd, grp); return error; } @@ -45,44 +44,44 @@ static int create_files(struct dentry * dir, int sysfs_create_group(struct kobject * kobj, const struct attribute_group * grp) { - struct dentry * dir; + struct sysfs_dirent *sd; int error; - BUG_ON(!kobj || !kobj->dentry); + BUG_ON(!kobj || !kobj->sd); if (grp->name) { - error = sysfs_create_subdir(kobj,grp->name,&dir); + error = sysfs_create_subdir(kobj, grp->name, &sd); if (error) return error; } else - dir = kobj->dentry; - dir = dget(dir); - if ((error = create_files(dir,grp))) { + sd = kobj->sd; + sysfs_get(sd); + error = create_files(sd, grp); + if (error) { if (grp->name) - sysfs_remove_subdir(dir); + sysfs_remove_subdir(sd); } - dput(dir); + sysfs_put(sd); return error; } void sysfs_remove_group(struct kobject * kobj, const struct attribute_group * grp) { - struct dentry * dir; + struct sysfs_dirent *dir_sd = kobj->sd; + struct sysfs_dirent *sd; if (grp->name) { - dir = lookup_one_len_kern(grp->name, kobj->dentry, - strlen(grp->name)); - BUG_ON(IS_ERR(dir)); - } - else - dir = dget(kobj->dentry); + sd = sysfs_get_dirent(dir_sd, grp->name); + BUG_ON(!sd); + } else + sd = sysfs_get(dir_sd); - remove_files(dir,grp); + remove_files(sd, grp); if (grp->name) - sysfs_remove_subdir(dir); - /* release the ref. taken in this routine */ - dput(dir); + sysfs_remove_subdir(sd); + + sysfs_put(sd); } diff --git a/fs/sysfs/inode.c b/fs/sysfs/inode.c index 5266eec15f6e..3756e152285a 100644 --- a/fs/sysfs/inode.c +++ b/fs/sysfs/inode.c @@ -133,187 +133,94 @@ static inline void set_inode_attr(struct inode * inode, struct iattr * iattr) */ static struct lock_class_key sysfs_inode_imutex_key; -struct inode * sysfs_new_inode(mode_t mode, struct sysfs_dirent * sd) +void sysfs_init_inode(struct sysfs_dirent *sd, struct inode *inode) { - struct inode * inode = new_inode(sysfs_sb); - if (inode) { - inode->i_blocks = 0; - inode->i_mapping->a_ops = &sysfs_aops; - inode->i_mapping->backing_dev_info = &sysfs_backing_dev_info; - inode->i_op = &sysfs_inode_operations; - inode->i_ino = sd->s_ino; - lockdep_set_class(&inode->i_mutex, &sysfs_inode_imutex_key); - - if (sd->s_iattr) { - /* sysfs_dirent has non-default attributes - * get them for the new inode from persistent copy - * in sysfs_dirent - */ - set_inode_attr(inode, sd->s_iattr); - } else - set_default_inode_attr(inode, mode); - } - return inode; -} - -int sysfs_create(struct dentry * dentry, int mode, int (*init)(struct inode *)) -{ - int error = 0; - struct inode * inode = NULL; - if (dentry) { - if (!dentry->d_inode) { - struct sysfs_dirent * sd = dentry->d_fsdata; - if ((inode = sysfs_new_inode(mode, sd))) { - if (dentry->d_parent && dentry->d_parent->d_inode) { - struct inode *p_inode = dentry->d_parent->d_inode; - p_inode->i_mtime = p_inode->i_ctime = CURRENT_TIME; - } - goto Proceed; - } - else - error = -ENOMEM; - } else - error = -EEXIST; - } else - error = -ENOENT; - goto Done; - - Proceed: - if (init) - error = init(inode); - if (!error) { - d_instantiate(dentry, inode); - if (S_ISDIR(mode)) - dget(dentry); /* pin only directory dentry in core */ + inode->i_blocks = 0; + inode->i_mapping->a_ops = &sysfs_aops; + inode->i_mapping->backing_dev_info = &sysfs_backing_dev_info; + inode->i_op = &sysfs_inode_operations; + inode->i_ino = sd->s_ino; + lockdep_set_class(&inode->i_mutex, &sysfs_inode_imutex_key); + + if (sd->s_iattr) { + /* sysfs_dirent has non-default attributes + * get them for the new inode from persistent copy + * in sysfs_dirent + */ + set_inode_attr(inode, sd->s_iattr); } else - iput(inode); - Done: - return error; + set_default_inode_attr(inode, sd->s_mode); } -/* - * Get the name for corresponding element represented by the given sysfs_dirent +/** + * sysfs_get_inode - get inode for sysfs_dirent + * @sd: sysfs_dirent to allocate inode for + * + * Get inode for @sd. If such inode doesn't exist, a new inode + * is allocated and basics are initialized. New inode is + * returned locked. + * + * LOCKING: + * Kernel thread context (may sleep). + * + * RETURNS: + * Pointer to allocated inode on success, NULL on failure. */ -const unsigned char * sysfs_get_name(struct sysfs_dirent *sd) +struct inode * sysfs_get_inode(struct sysfs_dirent *sd) { - struct attribute * attr; - struct bin_attribute * bin_attr; - struct sysfs_symlink * sl; - - BUG_ON(!sd || !sd->s_element); - - switch (sd->s_type) { - case SYSFS_DIR: - /* Always have a dentry so use that */ - return sd->s_dentry->d_name.name; - - case SYSFS_KOBJ_ATTR: - attr = sd->s_element; - return attr->name; - - case SYSFS_KOBJ_BIN_ATTR: - bin_attr = sd->s_element; - return bin_attr->attr.name; + struct inode *inode; - case SYSFS_KOBJ_LINK: - sl = sd->s_element; - return sl->link_name; - } - return NULL; -} + inode = iget_locked(sysfs_sb, sd->s_ino); + if (inode && (inode->i_state & I_NEW)) + sysfs_init_inode(sd, inode); -static inline void orphan_all_buffers(struct inode *node) -{ - struct sysfs_buffer_collection *set; - struct sysfs_buffer *buf; - - mutex_lock_nested(&node->i_mutex, I_MUTEX_CHILD); - set = node->i_private; - if (set) { - list_for_each_entry(buf, &set->associates, associates) { - down(&buf->sem); - buf->orphaned = 1; - up(&buf->sem); - } - } - mutex_unlock(&node->i_mutex); + return inode; } - -/* - * Unhashes the dentry corresponding to given sysfs_dirent - * Called with parent inode's i_mutex held. +/** + * sysfs_instantiate - instantiate dentry + * @dentry: dentry to be instantiated + * @inode: inode associated with @sd + * + * Unlock @inode if locked and instantiate @dentry with @inode. + * + * LOCKING: + * None. */ -void sysfs_drop_dentry(struct sysfs_dirent * sd, struct dentry * parent) +void sysfs_instantiate(struct dentry *dentry, struct inode *inode) { - struct dentry *dentry = NULL; - struct inode *inode; + BUG_ON(!dentry || dentry->d_inode); - /* We're not holding a reference to ->s_dentry dentry but the - * field will stay valid as long as sysfs_lock is held. - */ - spin_lock(&sysfs_lock); - spin_lock(&dcache_lock); - - /* dget dentry if it's still alive */ - if (sd->s_dentry && sd->s_dentry->d_inode) - dentry = dget_locked(sd->s_dentry); - - spin_unlock(&dcache_lock); - spin_unlock(&sysfs_lock); - - /* drop dentry */ - if (dentry) { - spin_lock(&dcache_lock); - spin_lock(&dentry->d_lock); - if (!d_unhashed(dentry) && dentry->d_inode) { - inode = dentry->d_inode; - spin_lock(&inode->i_lock); - __iget(inode); - spin_unlock(&inode->i_lock); - dget_locked(dentry); - __d_drop(dentry); - spin_unlock(&dentry->d_lock); - spin_unlock(&dcache_lock); - simple_unlink(parent->d_inode, dentry); - orphan_all_buffers(inode); - iput(inode); - } else { - spin_unlock(&dentry->d_lock); - spin_unlock(&dcache_lock); - } + if (inode->i_state & I_NEW) + unlock_new_inode(inode); - dput(dentry); - } + d_instantiate(dentry, inode); } -int sysfs_hash_and_remove(struct dentry * dir, const char * name) +int sysfs_hash_and_remove(struct sysfs_dirent *dir_sd, const char *name) { - struct sysfs_dirent * sd; - struct sysfs_dirent * parent_sd; - int found = 0; + struct sysfs_addrm_cxt acxt; + struct sysfs_dirent **pos, *sd; - if (!dir) + if (!dir_sd) return -ENOENT; - if (dir->d_inode == NULL) - /* no inode means this hasn't been made visible yet */ - return -ENOENT; + sysfs_addrm_start(&acxt, dir_sd); + + for (pos = &dir_sd->s_children; *pos; pos = &(*pos)->s_sibling) { + sd = *pos; - parent_sd = dir->d_fsdata; - mutex_lock_nested(&dir->d_inode->i_mutex, I_MUTEX_PARENT); - list_for_each_entry(sd, &parent_sd->s_children, s_sibling) { - if (!sd->s_element) + if (!sysfs_type(sd)) continue; - if (!strcmp(sysfs_get_name(sd), name)) { - list_del_init(&sd->s_sibling); - sysfs_drop_dentry(sd, dir); - sysfs_put(sd); - found = 1; + if (!strcmp(sd->s_name, name)) { + *pos = sd->s_sibling; + sd->s_sibling = NULL; + sysfs_remove_one(&acxt, sd); break; } } - mutex_unlock(&dir->d_inode->i_mutex); - return found ? 0 : -ENOENT; + if (sysfs_addrm_finish(&acxt)) + return 0; + return -ENOENT; } diff --git a/fs/sysfs/mount.c b/fs/sysfs/mount.c index 00ab9125d398..402cc356203c 100644 --- a/fs/sysfs/mount.c +++ b/fs/sysfs/mount.c @@ -19,28 +19,18 @@ struct vfsmount *sysfs_mount; struct super_block * sysfs_sb = NULL; struct kmem_cache *sysfs_dir_cachep; -static void sysfs_clear_inode(struct inode *inode); - static const struct super_operations sysfs_ops = { .statfs = simple_statfs, .drop_inode = sysfs_delete_inode, - .clear_inode = sysfs_clear_inode, }; -static struct sysfs_dirent sysfs_root = { - .s_sibling = LIST_HEAD_INIT(sysfs_root.s_sibling), - .s_children = LIST_HEAD_INIT(sysfs_root.s_children), - .s_element = NULL, - .s_type = SYSFS_ROOT, - .s_iattr = NULL, +struct sysfs_dirent sysfs_root = { + .s_count = ATOMIC_INIT(1), + .s_flags = SYSFS_ROOT, + .s_mode = S_IFDIR | S_IRWXU | S_IRUGO | S_IXUGO, .s_ino = 1, }; -static void sysfs_clear_inode(struct inode *inode) -{ - kfree(inode->i_private); -} - static int sysfs_fill_super(struct super_block *sb, void *data, int silent) { struct inode *inode; @@ -53,24 +43,26 @@ static int sysfs_fill_super(struct super_block *sb, void *data, int silent) sb->s_time_gran = 1; sysfs_sb = sb; - inode = sysfs_new_inode(S_IFDIR | S_IRWXU | S_IRUGO | S_IXUGO, - &sysfs_root); - if (inode) { - inode->i_op = &sysfs_dir_inode_operations; - inode->i_fop = &sysfs_dir_operations; - /* directory inodes start off with i_nlink == 2 (for "." entry) */ - inc_nlink(inode); - } else { + inode = new_inode(sysfs_sb); + if (!inode) { pr_debug("sysfs: could not get root inode\n"); return -ENOMEM; } + sysfs_init_inode(&sysfs_root, inode); + + inode->i_op = &sysfs_dir_inode_operations; + inode->i_fop = &sysfs_dir_operations; + /* directory inodes start off with i_nlink == 2 (for "." entry) */ + inc_nlink(inode); + root = d_alloc_root(inode); if (!root) { pr_debug("%s: could not get root dentry!\n",__FUNCTION__); iput(inode); return -ENOMEM; } + sysfs_root.s_dentry = root; root->d_fsdata = &sysfs_root; sb->s_root = root; return 0; diff --git a/fs/sysfs/symlink.c b/fs/sysfs/symlink.c index 7b9c5bfde920..2f86e0422290 100644 --- a/fs/sysfs/symlink.c +++ b/fs/sysfs/symlink.c @@ -11,71 +11,39 @@ #include "sysfs.h" -static int object_depth(struct kobject * kobj) +static int object_depth(struct sysfs_dirent *sd) { - struct kobject * p = kobj; int depth = 0; - do { depth++; } while ((p = p->parent)); + + for (; sd->s_parent; sd = sd->s_parent) + depth++; + return depth; } -static int object_path_length(struct kobject * kobj) +static int object_path_length(struct sysfs_dirent * sd) { - struct kobject * p = kobj; int length = 1; - do { - length += strlen(kobject_name(p)) + 1; - p = p->parent; - } while (p); + + for (; sd->s_parent; sd = sd->s_parent) + length += strlen(sd->s_name) + 1; + return length; } -static void fill_object_path(struct kobject * kobj, char * buffer, int length) +static void fill_object_path(struct sysfs_dirent *sd, char *buffer, int length) { - struct kobject * p; - --length; - for (p = kobj; p; p = p->parent) { - int cur = strlen(kobject_name(p)); + for (; sd->s_parent; sd = sd->s_parent) { + int cur = strlen(sd->s_name); /* back up enough to print this bus id with '/' */ length -= cur; - strncpy(buffer + length,kobject_name(p),cur); + strncpy(buffer + length, sd->s_name, cur); *(buffer + --length) = '/'; } } -static int sysfs_add_link(struct dentry * parent, const char * name, struct kobject * target) -{ - struct sysfs_dirent * parent_sd = parent->d_fsdata; - struct sysfs_symlink * sl; - int error = 0; - - error = -ENOMEM; - sl = kmalloc(sizeof(*sl), GFP_KERNEL); - if (!sl) - goto exit1; - - sl->link_name = kmalloc(strlen(name) + 1, GFP_KERNEL); - if (!sl->link_name) - goto exit2; - - strcpy(sl->link_name, name); - sl->target_kobj = kobject_get(target); - - error = sysfs_make_dirent(parent_sd, NULL, sl, S_IFLNK|S_IRWXUGO, - SYSFS_KOBJ_LINK); - if (!error) - return 0; - - kobject_put(target); - kfree(sl->link_name); -exit2: - kfree(sl); -exit1: - return error; -} - /** * sysfs_create_link - create symlink between two objects. * @kobj: object whose directory we're creating the link in. @@ -84,24 +52,57 @@ exit1: */ int sysfs_create_link(struct kobject * kobj, struct kobject * target, const char * name) { - struct dentry *dentry = NULL; - int error = -EEXIST; + struct sysfs_dirent *parent_sd = NULL; + struct sysfs_dirent *target_sd = NULL; + struct sysfs_dirent *sd = NULL; + struct sysfs_addrm_cxt acxt; + int error; BUG_ON(!name); if (!kobj) { if (sysfs_mount && sysfs_mount->mnt_sb) - dentry = sysfs_mount->mnt_sb->s_root; + parent_sd = sysfs_mount->mnt_sb->s_root->d_fsdata; } else - dentry = kobj->dentry; + parent_sd = kobj->sd; + + error = -EFAULT; + if (!parent_sd) + goto out_put; + + /* target->sd can go away beneath us but is protected with + * sysfs_assoc_lock. Fetch target_sd from it. + */ + spin_lock(&sysfs_assoc_lock); + if (target->sd) + target_sd = sysfs_get(target->sd); + spin_unlock(&sysfs_assoc_lock); + + error = -ENOENT; + if (!target_sd) + goto out_put; + + error = -ENOMEM; + sd = sysfs_new_dirent(name, S_IFLNK|S_IRWXUGO, SYSFS_KOBJ_LINK); + if (!sd) + goto out_put; + sd->s_elem.symlink.target_sd = target_sd; - if (!dentry) - return -EFAULT; + sysfs_addrm_start(&acxt, parent_sd); - mutex_lock(&dentry->d_inode->i_mutex); - if (!sysfs_dirent_exist(dentry->d_fsdata, name)) - error = sysfs_add_link(dentry, name, target); - mutex_unlock(&dentry->d_inode->i_mutex); + if (!sysfs_find_dirent(parent_sd, name)) { + sysfs_add_one(&acxt, sd); + sysfs_link_sibling(sd); + } + + if (sysfs_addrm_finish(&acxt)) + return 0; + + error = -EEXIST; + /* fall through */ + out_put: + sysfs_put(target_sd); + sysfs_put(sd); return error; } @@ -114,17 +115,17 @@ int sysfs_create_link(struct kobject * kobj, struct kobject * target, const char void sysfs_remove_link(struct kobject * kobj, const char * name) { - sysfs_hash_and_remove(kobj->dentry,name); + sysfs_hash_and_remove(kobj->sd, name); } -static int sysfs_get_target_path(struct kobject * kobj, struct kobject * target, - char *path) +static int sysfs_get_target_path(struct sysfs_dirent * parent_sd, + struct sysfs_dirent * target_sd, char *path) { char * s; int depth, size; - depth = object_depth(kobj); - size = object_path_length(target) + depth * 3 - 1; + depth = object_depth(parent_sd); + size = object_path_length(target_sd) + depth * 3 - 1; if (size > PATH_MAX) return -ENAMETOOLONG; @@ -133,7 +134,7 @@ static int sysfs_get_target_path(struct kobject * kobj, struct kobject * target, for (s = path; depth--; s += 3) strcpy(s,"../"); - fill_object_path(target, path, size); + fill_object_path(target_sd, path, size); pr_debug("%s: path = '%s'\n", __FUNCTION__, path); return 0; @@ -141,27 +142,16 @@ static int sysfs_get_target_path(struct kobject * kobj, struct kobject * target, static int sysfs_getlink(struct dentry *dentry, char * path) { - struct kobject *kobj, *target_kobj; - int error = 0; + struct sysfs_dirent *sd = dentry->d_fsdata; + struct sysfs_dirent *parent_sd = sd->s_parent; + struct sysfs_dirent *target_sd = sd->s_elem.symlink.target_sd; + int error; - kobj = sysfs_get_kobject(dentry->d_parent); - if (!kobj) - return -EINVAL; + mutex_lock(&sysfs_mutex); + error = sysfs_get_target_path(parent_sd, target_sd, path); + mutex_unlock(&sysfs_mutex); - target_kobj = sysfs_get_kobject(dentry); - if (!target_kobj) { - kobject_put(kobj); - return -EINVAL; - } - - down_read(&sysfs_rename_sem); - error = sysfs_get_target_path(kobj, target_kobj, path); - up_read(&sysfs_rename_sem); - - kobject_put(kobj); - kobject_put(target_kobj); return error; - } static void *sysfs_follow_link(struct dentry *dentry, struct nameidata *nd) diff --git a/fs/sysfs/sysfs.h b/fs/sysfs/sysfs.h index 502c949c402d..6a37f2386a8d 100644 --- a/fs/sysfs/sysfs.h +++ b/fs/sysfs/sysfs.h @@ -1,9 +1,40 @@ +struct sysfs_elem_dir { + struct kobject * kobj; +}; + +struct sysfs_elem_symlink { + struct sysfs_dirent * target_sd; +}; + +struct sysfs_elem_attr { + struct attribute * attr; +}; + +struct sysfs_elem_bin_attr { + struct bin_attribute * bin_attr; +}; + +/* + * As long as s_count reference is held, the sysfs_dirent itself is + * accessible. Dereferencing s_elem or any other outer entity + * requires s_active reference. + */ struct sysfs_dirent { atomic_t s_count; - struct list_head s_sibling; - struct list_head s_children; - void * s_element; - int s_type; + atomic_t s_active; + struct sysfs_dirent * s_parent; + struct sysfs_dirent * s_sibling; + struct sysfs_dirent * s_children; + const char * s_name; + + union { + struct sysfs_elem_dir dir; + struct sysfs_elem_symlink symlink; + struct sysfs_elem_attr attr; + struct sysfs_elem_bin_attr bin_attr; + } s_elem; + + unsigned int s_flags; umode_t s_mode; ino_t s_ino; struct dentry * s_dentry; @@ -11,30 +42,60 @@ struct sysfs_dirent { atomic_t s_event; }; +#define SD_DEACTIVATED_BIAS INT_MIN + +struct sysfs_addrm_cxt { + struct sysfs_dirent *parent_sd; + struct inode *parent_inode; + struct sysfs_dirent *removed; + int cnt; +}; + extern struct vfsmount * sysfs_mount; +extern struct sysfs_dirent sysfs_root; extern struct kmem_cache *sysfs_dir_cachep; -extern void sysfs_delete_inode(struct inode *inode); -extern struct inode * sysfs_new_inode(mode_t mode, struct sysfs_dirent *); -extern int sysfs_create(struct dentry *, int mode, int (*init)(struct inode *)); +extern struct dentry *sysfs_get_dentry(struct sysfs_dirent *sd); +extern void sysfs_link_sibling(struct sysfs_dirent *sd); +extern void sysfs_unlink_sibling(struct sysfs_dirent *sd); +extern struct sysfs_dirent *sysfs_get_active(struct sysfs_dirent *sd); +extern void sysfs_put_active(struct sysfs_dirent *sd); +extern struct sysfs_dirent *sysfs_get_active_two(struct sysfs_dirent *sd); +extern void sysfs_put_active_two(struct sysfs_dirent *sd); +extern void sysfs_addrm_start(struct sysfs_addrm_cxt *acxt, + struct sysfs_dirent *parent_sd); +extern void sysfs_add_one(struct sysfs_addrm_cxt *acxt, + struct sysfs_dirent *sd); +extern void sysfs_remove_one(struct sysfs_addrm_cxt *acxt, + struct sysfs_dirent *sd); +extern int sysfs_addrm_finish(struct sysfs_addrm_cxt *acxt); -extern int sysfs_dirent_exist(struct sysfs_dirent *, const unsigned char *); -extern int sysfs_make_dirent(struct sysfs_dirent *, struct dentry *, void *, - umode_t, int); - -extern int sysfs_add_file(struct dentry *, const struct attribute *, int); -extern int sysfs_hash_and_remove(struct dentry * dir, const char * name); +extern void sysfs_delete_inode(struct inode *inode); +extern void sysfs_init_inode(struct sysfs_dirent *sd, struct inode *inode); +extern struct inode * sysfs_get_inode(struct sysfs_dirent *sd); +extern void sysfs_instantiate(struct dentry *dentry, struct inode *inode); + +extern void release_sysfs_dirent(struct sysfs_dirent * sd); +extern struct sysfs_dirent *sysfs_find_dirent(struct sysfs_dirent *parent_sd, + const unsigned char *name); +extern struct sysfs_dirent *sysfs_get_dirent(struct sysfs_dirent *parent_sd, + const unsigned char *name); +extern struct sysfs_dirent *sysfs_new_dirent(const char *name, umode_t mode, + int type); + +extern int sysfs_add_file(struct sysfs_dirent *dir_sd, + const struct attribute *attr, int type); +extern int sysfs_hash_and_remove(struct sysfs_dirent *dir_sd, const char *name); extern struct sysfs_dirent *sysfs_find(struct sysfs_dirent *dir, const char * name); -extern int sysfs_create_subdir(struct kobject *, const char *, struct dentry **); -extern void sysfs_remove_subdir(struct dentry *); +extern int sysfs_create_subdir(struct kobject *kobj, const char *name, + struct sysfs_dirent **p_sd); +extern void sysfs_remove_subdir(struct sysfs_dirent *sd); -extern const unsigned char * sysfs_get_name(struct sysfs_dirent *sd); -extern void sysfs_drop_dentry(struct sysfs_dirent *sd, struct dentry *parent); extern int sysfs_setattr(struct dentry *dentry, struct iattr *iattr); -extern spinlock_t sysfs_lock; -extern struct rw_semaphore sysfs_rename_sem; +extern spinlock_t sysfs_assoc_lock; +extern struct mutex sysfs_mutex; extern struct super_block * sysfs_sb; extern const struct file_operations sysfs_dir_operations; extern const struct file_operations sysfs_file_operations; @@ -42,73 +103,9 @@ extern const struct file_operations bin_fops; extern const struct inode_operations sysfs_dir_inode_operations; extern const struct inode_operations sysfs_symlink_inode_operations; -struct sysfs_symlink { - char * link_name; - struct kobject * target_kobj; -}; - -struct sysfs_buffer { - struct list_head associates; - size_t count; - loff_t pos; - char * page; - struct sysfs_ops * ops; - struct semaphore sem; - int orphaned; - int needs_read_fill; - int event; -}; - -struct sysfs_buffer_collection { - struct list_head associates; -}; - -static inline struct kobject * to_kobj(struct dentry * dentry) -{ - struct sysfs_dirent * sd = dentry->d_fsdata; - return ((struct kobject *) sd->s_element); -} - -static inline struct attribute * to_attr(struct dentry * dentry) +static inline unsigned int sysfs_type(struct sysfs_dirent *sd) { - struct sysfs_dirent * sd = dentry->d_fsdata; - return ((struct attribute *) sd->s_element); -} - -static inline struct bin_attribute * to_bin_attr(struct dentry * dentry) -{ - struct sysfs_dirent * sd = dentry->d_fsdata; - return ((struct bin_attribute *) sd->s_element); -} - -static inline struct kobject *sysfs_get_kobject(struct dentry *dentry) -{ - struct kobject * kobj = NULL; - - spin_lock(&dcache_lock); - if (!d_unhashed(dentry)) { - struct sysfs_dirent * sd = dentry->d_fsdata; - if (sd->s_type & SYSFS_KOBJ_LINK) { - struct sysfs_symlink * sl = sd->s_element; - kobj = kobject_get(sl->target_kobj); - } else - kobj = kobject_get(sd->s_element); - } - spin_unlock(&dcache_lock); - - return kobj; -} - -static inline void release_sysfs_dirent(struct sysfs_dirent * sd) -{ - if (sd->s_type & SYSFS_KOBJ_LINK) { - struct sysfs_symlink * sl = sd->s_element; - kfree(sl->link_name); - kobject_put(sl->target_kobj); - kfree(sl); - } - kfree(sd->s_iattr); - kmem_cache_free(sysfs_dir_cachep, sd); + return sd->s_flags & SYSFS_TYPE_MASK; } static inline struct sysfs_dirent * sysfs_get(struct sysfs_dirent * sd) @@ -122,7 +119,7 @@ static inline struct sysfs_dirent * sysfs_get(struct sysfs_dirent * sd) static inline void sysfs_put(struct sysfs_dirent * sd) { - if (atomic_dec_and_test(&sd->s_count)) + if (sd && atomic_dec_and_test(&sd->s_count)) release_sysfs_dirent(sd); } |