diff options
Diffstat (limited to 'fs/nfs')
-rw-r--r-- | fs/nfs/Makefile | 1 | ||||
-rw-r--r-- | fs/nfs/client.c | 49 | ||||
-rw-r--r-- | fs/nfs/delegation.c | 6 | ||||
-rw-r--r-- | fs/nfs/dir.c | 263 | ||||
-rw-r--r-- | fs/nfs/direct.c | 8 | ||||
-rw-r--r-- | fs/nfs/file.c | 105 | ||||
-rw-r--r-- | fs/nfs/inode.c | 273 | ||||
-rw-r--r-- | fs/nfs/internal.h | 50 | ||||
-rw-r--r-- | fs/nfs/nfs2xdr.c | 20 | ||||
-rw-r--r-- | fs/nfs/nfs3acl.c | 2 | ||||
-rw-r--r-- | fs/nfs/nfs3proc.c | 17 | ||||
-rw-r--r-- | fs/nfs/nfs3xdr.c | 25 | ||||
-rw-r--r-- | fs/nfs/nfs4proc.c | 85 | ||||
-rw-r--r-- | fs/nfs/nfs4state.c | 2 | ||||
-rw-r--r-- | fs/nfs/nfs4xdr.c | 72 | ||||
-rw-r--r-- | fs/nfs/nfsroot.c | 3 | ||||
-rw-r--r-- | fs/nfs/proc.c | 5 | ||||
-rw-r--r-- | fs/nfs/read.c | 9 | ||||
-rw-r--r-- | fs/nfs/super.c | 393 | ||||
-rw-r--r-- | fs/nfs/unlink.c | 3 | ||||
-rw-r--r-- | fs/nfs/write.c | 199 |
21 files changed, 759 insertions, 831 deletions
diff --git a/fs/nfs/Makefile b/fs/nfs/Makefile index b55cb236cf7..df0f41e0988 100644 --- a/fs/nfs/Makefile +++ b/fs/nfs/Makefile @@ -16,4 +16,3 @@ nfs-$(CONFIG_NFS_V4) += nfs4proc.o nfs4xdr.o nfs4state.o nfs4renewd.o \ nfs4namespace.o nfs-$(CONFIG_NFS_DIRECTIO) += direct.o nfs-$(CONFIG_SYSCTL) += sysctl.o -nfs-objs := $(nfs-y) diff --git a/fs/nfs/client.c b/fs/nfs/client.c index a204484072f..a532ee12740 100644 --- a/fs/nfs/client.c +++ b/fs/nfs/client.c @@ -23,6 +23,8 @@ #include <linux/sunrpc/clnt.h> #include <linux/sunrpc/stats.h> #include <linux/sunrpc/metrics.h> +#include <linux/sunrpc/xprtsock.h> +#include <linux/sunrpc/xprtrdma.h> #include <linux/nfs_fs.h> #include <linux/nfs_mount.h> #include <linux/nfs4_mount.h> @@ -340,7 +342,8 @@ static void nfs_init_timeout_values(struct rpc_timeout *to, int proto, to->to_retries = 2; switch (proto) { - case IPPROTO_TCP: + case XPRT_TRANSPORT_TCP: + case XPRT_TRANSPORT_RDMA: if (!to->to_initval) to->to_initval = 60 * HZ; if (to->to_initval > NFS_MAX_TCP_TIMEOUT) @@ -349,7 +352,7 @@ static void nfs_init_timeout_values(struct rpc_timeout *to, int proto, to->to_maxval = to->to_initval + (to->to_increment * to->to_retries); to->to_exponential = 0; break; - case IPPROTO_UDP: + case XPRT_TRANSPORT_UDP: default: if (!to->to_initval) to->to_initval = 11 * HZ / 10; @@ -501,9 +504,9 @@ static int nfs_init_server_rpcclient(struct nfs_server *server, rpc_authflavor_t /* * Initialise an NFS2 or NFS3 client */ -static int nfs_init_client(struct nfs_client *clp, const struct nfs_mount_data *data) +static int nfs_init_client(struct nfs_client *clp, + const struct nfs_parsed_mount_data *data) { - int proto = (data->flags & NFS_MOUNT_TCP) ? IPPROTO_TCP : IPPROTO_UDP; int error; if (clp->cl_cons_state == NFS_CS_READY) { @@ -522,8 +525,8 @@ static int nfs_init_client(struct nfs_client *clp, const struct nfs_mount_data * * Create a client RPC handle for doing FSSTAT with UNIX auth only * - RFC 2623, sec 2.3.2 */ - error = nfs_create_rpc_client(clp, proto, data->timeo, data->retrans, - RPC_AUTH_UNIX, 0); + error = nfs_create_rpc_client(clp, data->nfs_server.protocol, + data->timeo, data->retrans, RPC_AUTH_UNIX, 0); if (error < 0) goto error; nfs_mark_client_ready(clp, NFS_CS_READY); @@ -538,7 +541,8 @@ error: /* * Create a version 2 or 3 client */ -static int nfs_init_server(struct nfs_server *server, const struct nfs_mount_data *data) +static int nfs_init_server(struct nfs_server *server, + const struct nfs_parsed_mount_data *data) { struct nfs_client *clp; int error, nfsvers = 2; @@ -551,7 +555,8 @@ static int nfs_init_server(struct nfs_server *server, const struct nfs_mount_dat #endif /* Allocate or find a client reference we can use */ - clp = nfs_get_client(data->hostname, &data->addr, nfsvers); + clp = nfs_get_client(data->nfs_server.hostname, + &data->nfs_server.address, nfsvers); if (IS_ERR(clp)) { dprintk("<-- nfs_init_server() = error %ld\n", PTR_ERR(clp)); return PTR_ERR(clp); @@ -581,7 +586,7 @@ static int nfs_init_server(struct nfs_server *server, const struct nfs_mount_dat if (error < 0) goto error; - error = nfs_init_server_rpcclient(server, data->pseudoflavor); + error = nfs_init_server_rpcclient(server, data->auth_flavors[0]); if (error < 0) goto error; @@ -760,7 +765,7 @@ void nfs_free_server(struct nfs_server *server) * Create a version 2 or 3 volume record * - keyed on server and FSID */ -struct nfs_server *nfs_create_server(const struct nfs_mount_data *data, +struct nfs_server *nfs_create_server(const struct nfs_parsed_mount_data *data, struct nfs_fh *mntfh) { struct nfs_server *server; @@ -906,7 +911,7 @@ error: * Create a version 4 volume record */ static int nfs4_init_server(struct nfs_server *server, - const struct nfs4_mount_data *data, rpc_authflavor_t authflavour) + const struct nfs_parsed_mount_data *data) { int error; @@ -926,7 +931,7 @@ static int nfs4_init_server(struct nfs_server *server, server->acdirmin = data->acdirmin * HZ; server->acdirmax = data->acdirmax * HZ; - error = nfs_init_server_rpcclient(server, authflavour); + error = nfs_init_server_rpcclient(server, data->auth_flavors[0]); /* Done */ dprintk("<-- nfs4_init_server() = %d\n", error); @@ -937,12 +942,7 @@ static int nfs4_init_server(struct nfs_server *server, * Create a version 4 volume record * - keyed on server and FSID */ -struct nfs_server *nfs4_create_server(const struct nfs4_mount_data *data, - const char *hostname, - const struct sockaddr_in *addr, - const char *mntpath, - const char *ip_addr, - rpc_authflavor_t authflavour, +struct nfs_server *nfs4_create_server(const struct nfs_parsed_mount_data *data, struct nfs_fh *mntfh) { struct nfs_fattr fattr; @@ -956,13 +956,18 @@ struct nfs_server *nfs4_create_server(const struct nfs4_mount_data *data, return ERR_PTR(-ENOMEM); /* Get a client record */ - error = nfs4_set_client(server, hostname, addr, ip_addr, authflavour, - data->proto, data->timeo, data->retrans); + error = nfs4_set_client(server, + data->nfs_server.hostname, + &data->nfs_server.address, + data->client_address, + data->auth_flavors[0], + data->nfs_server.protocol, + data->timeo, data->retrans); if (error < 0) goto error; /* set up the general RPC client */ - error = nfs4_init_server(server, data, authflavour); + error = nfs4_init_server(server, data); if (error < 0) goto error; @@ -971,7 +976,7 @@ struct nfs_server *nfs4_create_server(const struct nfs4_mount_data *data, BUG_ON(!server->nfs_client->rpc_ops->file_inode_ops); /* Probe the root fh to retrieve its FSID */ - error = nfs4_path_walk(server, mntfh, mntpath); + error = nfs4_path_walk(server, mntfh, data->nfs_server.export_path); if (error < 0) goto error; diff --git a/fs/nfs/delegation.c b/fs/nfs/delegation.c index c55a761c22b..af8b235d405 100644 --- a/fs/nfs/delegation.c +++ b/fs/nfs/delegation.c @@ -52,7 +52,7 @@ static int nfs_delegation_claim_locks(struct nfs_open_context *ctx, struct nfs4_ for (fl = inode->i_flock; fl != 0; fl = fl->fl_next) { if (!(fl->fl_flags & (FL_POSIX|FL_FLOCK))) continue; - if ((struct nfs_open_context *)fl->fl_file->private_data != ctx) + if (nfs_file_open_context(fl->fl_file) != ctx) continue; status = nfs4_lock_delegation_recall(state, fl); if (status >= 0) @@ -109,6 +109,7 @@ again: void nfs_inode_reclaim_delegation(struct inode *inode, struct rpc_cred *cred, struct nfs_openres *res) { struct nfs_delegation *delegation = NFS_I(inode)->delegation; + struct rpc_cred *oldcred; if (delegation == NULL) return; @@ -116,11 +117,12 @@ void nfs_inode_reclaim_delegation(struct inode *inode, struct rpc_cred *cred, st sizeof(delegation->stateid.data)); delegation->type = res->delegation_type; delegation->maxsize = res->maxsize; - put_rpccred(cred); + oldcred = delegation->cred; delegation->cred = get_rpccred(cred); delegation->flags &= ~NFS_DELEGATION_NEED_RECLAIM; NFS_I(inode)->delegation_state = delegation->type; smp_wmb(); + put_rpccred(oldcred); } /* diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c index e4a04d16b8b..8ec7fbd8240 100644 --- a/fs/nfs/dir.c +++ b/fs/nfs/dir.c @@ -200,9 +200,6 @@ int nfs_readdir_filler(nfs_readdir_descriptor_t *desc, struct page *page) desc->timestamp = timestamp; desc->timestamp_valid = 1; SetPageUptodate(page); - spin_lock(&inode->i_lock); - NFS_I(inode)->cache_validity |= NFS_INO_INVALID_ATIME; - spin_unlock(&inode->i_lock); /* Ensure consistent page alignment of the data. * Note: assumes we have exclusive access to this mapping either * through inode->i_mutex or some other mechanism. @@ -214,9 +211,7 @@ int nfs_readdir_filler(nfs_readdir_descriptor_t *desc, struct page *page) unlock_page(page); return 0; error: - SetPageError(page); unlock_page(page); - nfs_zap_caches(inode); desc->error = error; return -EIO; } @@ -407,7 +402,7 @@ int nfs_do_filldir(nfs_readdir_descriptor_t *desc, void *dirent, struct file *file = desc->file; struct nfs_entry *entry = desc->entry; struct dentry *dentry = NULL; - unsigned long fileid; + u64 fileid; int loop_count = 0, res; @@ -418,7 +413,7 @@ int nfs_do_filldir(nfs_readdir_descriptor_t *desc, void *dirent, unsigned d_type = DT_UNKNOWN; /* Note: entry->prev_cookie contains the cookie for * retrieving the current dirent on the server */ - fileid = nfs_fileid_to_ino_t(entry->ino); + fileid = entry->ino; /* Get a dentry if we have one */ if (dentry != NULL) @@ -428,11 +423,12 @@ int nfs_do_filldir(nfs_readdir_descriptor_t *desc, void *dirent, /* Use readdirplus info */ if (dentry != NULL && dentry->d_inode != NULL) { d_type = dt_type(dentry->d_inode); - fileid = dentry->d_inode->i_ino; + fileid = NFS_FILEID(dentry->d_inode); } res = filldir(dirent, entry->name, entry->len, - file->f_pos, fileid, d_type); + file->f_pos, nfs_compat_user_ino64(fileid), + d_type); if (res < 0) break; file->f_pos++; @@ -490,9 +486,6 @@ int uncached_readdir(nfs_readdir_descriptor_t *desc, void *dirent, page, NFS_SERVER(inode)->dtsize, desc->plus); - spin_lock(&inode->i_lock); - NFS_I(inode)->cache_validity |= NFS_INO_INVALID_ATIME; - spin_unlock(&inode->i_lock); desc->page = page; desc->ptr = kmap(page); /* matching kunmap in nfs_do_filldir */ if (desc->error >= 0) { @@ -558,7 +551,7 @@ static int nfs_readdir(struct file *filp, void *dirent, filldir_t filldir) memset(desc, 0, sizeof(*desc)); desc->file = filp; - desc->dir_cookie = &((struct nfs_open_context *)filp->private_data)->dir_cookie; + desc->dir_cookie = &nfs_file_open_context(filp)->dir_cookie; desc->decode = NFS_PROTO(inode)->decode_dirent; desc->plus = NFS_USE_READDIRPLUS(inode); @@ -623,7 +616,7 @@ static loff_t nfs_llseek_dir(struct file *filp, loff_t offset, int origin) } if (offset != filp->f_pos) { filp->f_pos = offset; - ((struct nfs_open_context *)filp->private_data)->dir_cookie = 0; + nfs_file_open_context(filp)->dir_cookie = 0; } out: mutex_unlock(&filp->f_path.dentry->d_inode->i_mutex); @@ -650,36 +643,18 @@ static int nfs_fsync_dir(struct file *filp, struct dentry *dentry, int datasync) */ static int nfs_check_verifier(struct inode *dir, struct dentry *dentry) { - unsigned long verf; - if (IS_ROOT(dentry)) return 1; - verf = dentry->d_time; - if (nfs_caches_unstable(dir) - || verf != NFS_I(dir)->cache_change_attribute) + if (!nfs_verify_change_attribute(dir, dentry->d_time)) + return 0; + /* Revalidate nfsi->cache_change_attribute before we declare a match */ + if (nfs_revalidate_inode(NFS_SERVER(dir), dir) < 0) + return 0; + if (!nfs_verify_change_attribute(dir, dentry->d_time)) return 0; return 1; } -static inline void nfs_set_verifier(struct dentry * dentry, unsigned long verf) -{ - dentry->d_time = verf; -} - -static void nfs_refresh_verifier(struct dentry * dentry, unsigned long verf) -{ - nfs_set_verifier(dentry, verf); -} - -/* - * Whenever an NFS operation succeeds, we know that the dentry - * is valid, so we update the revalidation timestamp. - */ -static inline void nfs_renew_times(struct dentry * dentry) -{ - dentry->d_time = jiffies; -} - /* * Return the intent data that applies to this particular path component * @@ -695,6 +670,19 @@ static inline unsigned int nfs_lookup_check_intent(struct nameidata *nd, unsigne } /* + * Use intent information to check whether or not we're going to do + * an O_EXCL create using this path component. + */ +static int nfs_is_exclusive_create(struct inode *dir, struct nameidata *nd) +{ + if (NFS_PROTO(dir)->version == 2) + return 0; + if (nd == NULL || nfs_lookup_check_intent(nd, LOOKUP_CREATE) == 0) + return 0; + return (nd->intent.open.flags & O_EXCL) != 0; +} + +/* * Inode and filehandle revalidation for lookups. * * We force revalidation in the cases where the VFS sets LOOKUP_REVAL, @@ -717,6 +705,7 @@ int nfs_lookup_verify_inode(struct inode *inode, struct nameidata *nd) (S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode))) goto out_force; + return 0; } return nfs_revalidate_inode(server, inode); out_force: @@ -759,7 +748,6 @@ static int nfs_lookup_revalidate(struct dentry * dentry, struct nameidata *nd) int error; struct nfs_fh fhandle; struct nfs_fattr fattr; - unsigned long verifier; parent = dget_parent(dentry); lock_kernel(); @@ -767,10 +755,6 @@ static int nfs_lookup_revalidate(struct dentry * dentry, struct nameidata *nd) nfs_inc_stats(dir, NFSIOS_DENTRYREVALIDATE); inode = dentry->d_inode; - /* Revalidate parent directory attribute cache */ - if (nfs_revalidate_inode(NFS_SERVER(dir), dir) < 0) - goto out_zap_parent; - if (!inode) { if (nfs_neg_need_reval(dir, dentry, nd)) goto out_bad; @@ -785,7 +769,7 @@ static int nfs_lookup_revalidate(struct dentry * dentry, struct nameidata *nd) } /* Force a full look up iff the parent directory has changed */ - if (nfs_check_verifier(dir, dentry)) { + if (!nfs_is_exclusive_create(dir, nd) && nfs_check_verifier(dir, dentry)) { if (nfs_lookup_verify_inode(inode, nd)) goto out_zap_parent; goto out_valid; @@ -794,7 +778,6 @@ static int nfs_lookup_revalidate(struct dentry * dentry, struct nameidata *nd) if (NFS_STALE(inode)) goto out_bad; - verifier = nfs_save_change_attribute(dir); error = NFS_PROTO(dir)->lookup(dir, &dentry->d_name, &fhandle, &fattr); if (error) goto out_bad; @@ -803,8 +786,7 @@ static int nfs_lookup_revalidate(struct dentry * dentry, struct nameidata *nd) if ((error = nfs_refresh_inode(inode, &fattr)) != 0) goto out_bad; - nfs_renew_times(dentry); - nfs_refresh_verifier(dentry, verifier); + nfs_set_verifier(dentry, nfs_save_change_attribute(dir)); out_valid: unlock_kernel(); dput(parent); @@ -815,7 +797,7 @@ static int nfs_lookup_revalidate(struct dentry * dentry, struct nameidata *nd) out_zap_parent: nfs_zap_caches(dir); out_bad: - NFS_CACHEINV(dir); + nfs_mark_for_revalidate(dir); if (inode && S_ISDIR(inode->i_mode)) { /* Purge readdir caches. */ nfs_zap_caches(inode); @@ -872,8 +854,6 @@ static void nfs_dentry_iput(struct dentry *dentry, struct inode *inode) nfs_complete_unlink(dentry, inode); unlock_kernel(); } - /* When creating a negative dentry, we want to renew d_time */ - nfs_renew_times(dentry); iput(inode); } @@ -883,30 +863,6 @@ struct dentry_operations nfs_dentry_operations = { .d_iput = nfs_dentry_iput, }; -/* - * Use intent information to check whether or not we're going to do - * an O_EXCL create using this path component. - */ -static inline -int nfs_is_exclusive_create(struct inode *dir, struct nameidata *nd) -{ - if (NFS_PROTO(dir)->version == 2) - return 0; - if (nd == NULL || nfs_lookup_check_intent(nd, LOOKUP_CREATE) == 0) - return 0; - return (nd->intent.open.flags & O_EXCL) != 0; -} - -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 using the parent directory */ - return __nfs_revalidate_inode(server, dir); - return 0; -} - static struct dentry *nfs_lookup(struct inode *dir, struct dentry * dentry, struct nameidata *nd) { struct dentry *res; @@ -945,11 +901,6 @@ static struct dentry *nfs_lookup(struct inode *dir, struct dentry * dentry, stru res = ERR_PTR(error); goto out_unlock; } - error = nfs_reval_fsid(dir, &fattr); - if (error < 0) { - res = ERR_PTR(error); - goto out_unlock; - } inode = nfs_fhget(dentry->d_sb, &fhandle, &fattr); res = (struct dentry *)inode; if (IS_ERR(res)) @@ -958,17 +909,10 @@ static struct dentry *nfs_lookup(struct inode *dir, struct dentry * dentry, stru no_entry: res = d_materialise_unique(dentry, inode); if (res != NULL) { - struct dentry *parent; if (IS_ERR(res)) goto out_unlock; - /* Was a directory renamed! */ - parent = dget_parent(res); - if (!IS_ROOT(parent)) - nfs_mark_for_revalidate(parent->d_inode); - dput(parent); dentry = res; } - nfs_renew_times(dentry); nfs_set_verifier(dentry, nfs_save_change_attribute(dir)); out_unlock: unlock_kernel(); @@ -1020,28 +964,16 @@ static struct dentry *nfs_atomic_lookup(struct inode *dir, struct dentry *dentry } dentry->d_op = NFS_PROTO(dir)->dentry_ops; - /* Let vfs_create() deal with O_EXCL */ + /* Let vfs_create() deal with O_EXCL. Instantiate, but don't hash + * the dentry. */ if (nd->intent.open.flags & O_EXCL) { - d_add(dentry, NULL); + d_instantiate(dentry, NULL); goto out; } /* Open the file on the server */ lock_kernel(); - /* Revalidate parent directory attribute cache */ - error = nfs_revalidate_inode(NFS_SERVER(dir), dir); - if (error < 0) { - res = ERR_PTR(error); - unlock_kernel(); - goto out; - } - - if (nd->intent.open.flags & O_CREAT) { - nfs_begin_data_update(dir); - res = nfs4_atomic_open(dir, dentry, nd); - nfs_end_data_update(dir); - } else - res = nfs4_atomic_open(dir, dentry, nd); + res = nfs4_atomic_open(dir, dentry, nd); unlock_kernel(); if (IS_ERR(res)) { error = PTR_ERR(res); @@ -1063,8 +995,6 @@ static struct dentry *nfs_atomic_lookup(struct inode *dir, struct dentry *dentry } } else if (res != NULL) dentry = res; - nfs_renew_times(dentry); - nfs_set_verifier(dentry, nfs_save_change_attribute(dir)); out: return res; no_open: @@ -1076,7 +1006,6 @@ static int nfs_open_revalidate(struct dentry *dentry, struct nameidata *nd) struct dentry *parent = NULL; struct inode *inode = dentry->d_inode; struct inode *dir; - unsigned long verifier; int openflags, ret = 0; parent = dget_parent(dentry); @@ -1086,8 +1015,12 @@ static int nfs_open_revalidate(struct dentry *dentry, struct nameidata *nd) /* We can't create new files in nfs_open_revalidate(), so we * optimize away revalidation of negative dentries. */ - if (inode == NULL) + if (inode == NULL) { + if (!nfs_neg_need_reval(dir, dentry, nd)) + ret = 1; goto out; + } + /* NFS only supports OPEN on regular files */ if (!S_ISREG(inode->i_mode)) goto no_open; @@ -1104,10 +1037,7 @@ static int nfs_open_revalidate(struct dentry *dentry, struct nameidata *nd) * change attribute *before* we do the RPC call. */ lock_kernel(); - verifier = nfs_save_change_attribute(dir); ret = nfs4_open_revalidate(dir, dentry, openflags, nd); - if (!ret) - nfs_refresh_verifier(dentry, verifier); unlock_kernel(); out: dput(parent); @@ -1133,6 +1063,7 @@ static struct dentry *nfs_readdir_lookup(nfs_readdir_descriptor_t *desc) .len = entry->len, }; struct inode *inode; + unsigned long verf = nfs_save_change_attribute(dir); switch (name.len) { case 2: @@ -1143,6 +1074,14 @@ static struct dentry *nfs_readdir_lookup(nfs_readdir_descriptor_t *desc) if (name.name[0] == '.') return dget(parent); } + + spin_lock(&dir->i_lock); + if (NFS_I(dir)->cache_validity & NFS_INO_INVALID_DATA) { + spin_unlock(&dir->i_lock); + return NULL; + } + spin_unlock(&dir->i_lock); + name.hash = full_name_hash(name.name, name.len); dentry = d_lookup(parent, &name); if (dentry != NULL) { @@ -1183,12 +1122,8 @@ static struct dentry *nfs_readdir_lookup(nfs_readdir_descriptor_t *desc) dentry = alias; } - nfs_renew_times(dentry); - nfs_set_verifier(dentry, nfs_save_change_attribute(dir)); - return dentry; out_renew: - nfs_renew_times(dentry); - nfs_refresh_verifier(dentry, nfs_save_change_attribute(dir)); + nfs_set_verifier(dentry, verf); return dentry; } @@ -1198,32 +1133,40 @@ out_renew: int nfs_instantiate(struct dentry *dentry, struct nfs_fh *fhandle, struct nfs_fattr *fattr) { + struct dentry *parent = dget_parent(dentry); + struct inode *dir = parent->d_inode; struct inode *inode; int error = -EACCES; + d_drop(dentry); + /* We may have been initialized further down */ if (dentry->d_inode) - return 0; + goto out; if (fhandle->size == 0) { - struct inode *dir = dentry->d_parent->d_inode; error = NFS_PROTO(dir)->lookup(dir, &dentry->d_name, fhandle, fattr); if (error) - return error; + goto out_error; } + nfs_set_verifier(dentry, nfs_save_change_attribute(dir)); if (!(fattr->valid & NFS_ATTR_FATTR)) { struct nfs_server *server = NFS_SB(dentry->d_sb); error = server->nfs_client->rpc_ops->getattr(server, fhandle, fattr); if (error < 0) - return error; + goto out_error; } inode = nfs_fhget(dentry->d_sb, fhandle, fattr); error = PTR_ERR(inode); if (IS_ERR(inode)) - return error; - d_instantiate(dentry, inode); - if (d_unhashed(dentry)) - d_rehash(dentry); + goto out_error; + d_add(dentry, inode); +out: + dput(parent); return 0; +out_error: + nfs_mark_for_revalidate(dir); + dput(parent); + return error; } /* @@ -1249,13 +1192,9 @@ static int nfs_create(struct inode *dir, struct dentry *dentry, int mode, open_flags = nd->intent.open.flags; lock_kernel(); - nfs_begin_data_update(dir); error = NFS_PROTO(dir)->create(dir, dentry, &attr, open_flags, nd); - nfs_end_data_update(dir); if (error != 0) goto out_err; - nfs_renew_times(dentry); - nfs_set_verifier(dentry, nfs_save_change_attribute(dir)); unlock_kernel(); return 0; out_err: @@ -1283,13 +1222,9 @@ nfs_mknod(struct inode *dir, struct dentry *dentry, int mode, dev_t rdev) attr.ia_valid = ATTR_MODE; lock_kernel(); - nfs_begin_data_update(dir); status = NFS_PROTO(dir)->mknod(dir, dentry, &attr, rdev); - nfs_end_data_update(dir); if (status != 0) goto out_err; - nfs_renew_times(dentry); - nfs_set_verifier(dentry, nfs_save_change_attribute(dir)); unlock_kernel(); return 0; out_err: @@ -1313,13 +1248,9 @@ static int nfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) attr.ia_mode = mode | S_IFDIR; lock_kernel(); - nfs_begin_data_update(dir); error = NFS_PROTO(dir)->mkdir(dir, dentry, &attr); - nfs_end_data_update(dir); if (error != 0) goto out_err; - nfs_renew_times(dentry); - nfs_set_verifier(dentry, nfs_save_change_attribute(dir)); unlock_kernel(); return 0; out_err: @@ -1336,12 +1267,10 @@ static int nfs_rmdir(struct inode *dir, struct dentry *dentry) dir->i_sb->s_id, dir->i_ino, dentry->d_name.name); lock_kernel(); - nfs_begin_data_update(dir); error = NFS_PROTO(dir)->rmdir(dir, &dentry->d_name); /* Ensure the VFS deletes this inode */ if (error == 0 && dentry->d_inode != NULL) clear_nlink(dentry->d_inode); - nfs_end_data_update(dir); unlock_kernel(); return error; @@ -1350,9 +1279,9 @@ static int nfs_rmdir(struct inode *dir, struct dentry *dentry) static int nfs_sillyrename(struct inode *dir, struct dentry *dentry) { static unsigned int sillycounter; - const int i_inosize = sizeof(dir->i_ino)*2; + const int fileidsize = sizeof(NFS_FILEID(dentry->d_inode))*2; const int countersize = sizeof(sillycounter)*2; - const int slen = sizeof(".nfs") + i_inosize + countersize - 1; + const int slen = sizeof(".nfs")+fileidsize+countersize-1; char silly[slen+1]; struct qstr qsilly; struct dentry *sdentry; @@ -1370,8 +1299,9 @@ static int nfs_sillyrename(struct inode *dir, struct dentry *dentry) if (dentry->d_flags & DCACHE_NFSFS_RENAMED) goto out; - sprintf(silly, ".nfs%*.*lx", - i_inosize, i_inosize, dentry->d_inode->i_ino); + sprintf(silly, ".nfs%*.*Lx", + fileidsize, fileidsize, + (unsigned long long)NFS_FILEID(dentry->d_inode)); /* Return delegation in anticipation of the rename */ nfs_inode_return_delegation(dentry->d_inode); @@ -1398,19 +1328,14 @@ static int nfs_sillyrename(struct inode *dir, struct dentry *dentry) qsilly.name = silly; qsilly.len = strlen(silly); - nfs_begin_data_update(dir); if (dentry->d_inode) { - nfs_begin_data_update(dentry->d_inode); error = NFS_PROTO(dir)->rename(dir, &dentry->d_name, dir, &qsilly); nfs_mark_for_revalidate(dentry->d_inode); - nfs_end_data_update(dentry->d_inode); } else error = NFS_PROTO(dir)->rename(dir, &dentry->d_name, dir, &qsilly); - nfs_end_data_update(dir); if (!error) { - nfs_renew_times(dentry); nfs_set_verifier(dentry, nfs_save_change_attribute(dir)); d_move(dentry, sdentry); error = nfs_async_unlink(dir, dentry); @@ -1443,19 +1368,15 @@ static int nfs_safe_remove(struct dentry *dentry) goto out; } - nfs_begin_data_update(dir); if (inode != NULL) { nfs_inode_return_delegation(inode); - nfs_begin_data_update(inode); error = NFS_PROTO(dir)->remove(dir, &dentry->d_name); /* The VFS may want to delete this inode */ if (error == 0) drop_nlink(inode); nfs_mark_for_revalidate(inode); - nfs_end_data_update(inode); } else error = NFS_PROTO(dir)->remove(dir, &dentry->d_name); - nfs_end_data_update(dir); out: return error; } @@ -1493,7 +1414,6 @@ static int nfs_unlink(struct inode *dir, struct dentry *dentry) spin_unlock(&dcache_lock); error = nfs_safe_remove(dentry); if (!error) { - nfs_renew_times(dentry); nfs_set_verifier(dentry, nfs_save_change_attribute(dir)); } else if (need_rehash) d_rehash(dentry); @@ -1548,9 +1468,7 @@ static int nfs_symlink(struct inode *dir, struct dentry *dentry, const char *sym memset(kaddr + pathlen, 0, PAGE_SIZE - pathlen); kunmap_atomic(kaddr, KM_USER0); - nfs_begin_data_update(dir); error = NFS_PROTO(dir)->symlink(dir, dentry, page, pathlen, &attr); - nfs_end_data_update(dir); if (error != 0) { dfprintk(VFS, "NFS: symlink(%s/%ld, %s, %s) error %d\n", dir->i_sb->s_id, dir->i_ino, @@ -1590,15 +1508,12 @@ nfs_link(struct dentry *old_dentry, struct inode *dir, struct dentry *dentry) dentry->d_parent->d_name.name, dentry->d_name.name); lock_kernel(); - nfs_begin_data_update(dir); - nfs_begin_data_update(inode); + d_drop(dentry); error = NFS_PROTO(dir)->link(inode, dir, &dentry->d_name); if (error == 0) { atomic_inc(&inode->i_count); - d_instantiate(dentry, inode); + d_add(dentry, inode); } - nfs_end_data_update(inode); - nfs_end_data_update(dir); unlock_kernel(); return error; } @@ -1701,22 +1616,16 @@ go_ahead: d_delete(new_dentry); } - nfs_begin_data_update(old_dir); - nfs_begin_data_update(new_dir); - nfs_begin_data_update(old_inode); error = NFS_PROTO(old_dir)->rename(old_dir, &old_dentry->d_name, new_dir, &new_dentry->d_name); nfs_mark_for_revalidate(old_inode); - nfs_end_data_update(old_inode); - nfs_end_data_update(new_dir); - nfs_end_data_update(old_dir); out: if (rehash) d_rehash(rehash); if (!error) { d_move(old_dentry, new_dentry); - nfs_renew_times(new_dentry); - nfs_refresh_verifier(new_dentry, nfs_save_change_attribute(new_dir)); + nfs_set_verifier(new_dentry, + nfs_save_change_attribute(new_dir)); } /* new dentry created? */ @@ -1842,7 +1751,7 @@ static struct nfs_access_entry *nfs_access_search_rbtree(struct inode *inode, st return NULL; } -int nfs_access_get_cached(struct inode *inode, struct rpc_cred *cred, struct nfs_access_entry *res) +static int nfs_access_get_cached(struct inode *inode, struct rpc_cred *cred, struct nfs_access_entry *res) { struct nfs_inode *nfsi = NFS_I(inode); struct nfs_access_entry *cache; @@ -1854,7 +1763,7 @@ int nfs_access_get_cached(struct inode *inode, struct rpc_cred *cred, struct nfs cache = nfs_access_search_rbtree(inode, cred); if (cache == NULL) goto out; - if (time_after(jiffies, cache->jiffies + NFS_ATTRTIMEO(inode))) + if (!time_in_range(jiffies, cache->jiffies, cache->jiffies + nfsi->attrtimeo)) goto out_stale; res->jiffies = cache->jiffies; res->cred = cache->cred; @@ -1909,7 +1818,7 @@ found: nfs_access_free_entry(entry); } -void nfs_access_add_cache(struct inode *inode, struct nfs_access_entry *set) +static void nfs_access_add_cache(struct inode *inode, struct nfs_access_entry *set) { struct nfs_access_entry *cache = kmalloc(sizeof(*cache), GFP_KERNEL); if (cache == NULL) @@ -1957,6 +1866,24 @@ out: return -EACCES; } +static int nfs_open_permission_mask(int openflags) +{ + int mask = 0; + + if (openflags & FMODE_READ) + mask |= MAY_READ; + if (openflags & FMODE_WRITE) + mask |= MAY_WRITE; + if (openflags & FMODE_EXEC) + mask |= MAY_EXEC; + return mask; +} + +int nfs_may_open(struct inode *inode, struct rpc_cred *cred, int openflags) +{ + return nfs_do_access(inode, cred, nfs_open_permission_mask(openflags)); +} + int nfs_permission(struct inode *inode, int mask, struct nameidata *nd) { struct rpc_cred *cred; diff --git a/fs/nfs/direct.c b/fs/nfs/direct.c index fcf4d384610..32fe97211ee 100644 --- a/fs/nfs/direct.c +++ b/fs/nfs/direct.c @@ -368,7 +368,7 @@ static ssize_t nfs_direct_read(struct kiocb *iocb, unsigned long user_addr, size return -ENOMEM; dreq->inode = inode; - dreq->ctx = get_nfs_open_context((struct nfs_open_context *)iocb->ki_filp->private_data); + dreq->ctx = get_nfs_open_context(nfs_file_open_context(iocb->ki_filp)); if (!is_sync_kiocb(iocb)) dreq->iocb = iocb; @@ -510,7 +510,6 @@ static void nfs_direct_write_complete(struct nfs_direct_req *dreq, struct inode nfs_direct_write_reschedule(dreq); break; default: - nfs_end_data_update(inode); if (dreq->commit_data != NULL) nfs_commit_free(dreq->commit_data); nfs_direct_free_writedata(dreq); @@ -533,7 +532,6 @@ static inline void nfs_alloc_commit_data(struct nfs_direct_req *dreq) static void nfs_direct_write_complete(struct nfs_direct_req *dreq, struct inode *inode) { - nfs_end_data_update(inode); nfs_direct_free_writedata(dreq); nfs_zap_mapping(inode, inode->i_mapping); nfs_direct_complete(dreq); @@ -718,14 +716,12 @@ static ssize_t nfs_direct_write(struct kiocb *iocb, unsigned long user_addr, siz sync = FLUSH_STABLE; dreq->inode = inode; - dreq->ctx = get_nfs_open_context((struct nfs_open_context *)iocb->ki_filp->private_data); + dreq->ctx = get_nfs_open_context(nfs_file_open_context(iocb->ki_filp)); if (!is_sync_kiocb(iocb)) dreq->iocb = iocb; nfs_add_stats(inode, NFSIOS_DIRECTWRITTENBYTES, count); - nfs_begin_data_update(inode); - rpc_clnt_sigmask(clnt, &oldset); result = nfs_direct_write_schedule(dreq, user_addr, count, pos, sync); if (!result) diff --git a/fs/nfs/file.c b/fs/nfs/file.c index 9c98ccbf9de..08c7c7387fc 100644 --- a/fs/nfs/file.c +++ b/fs/nfs/file.c @@ -33,6 +33,7 @@ #include <asm/system.h> #include "delegation.h" +#include "internal.h" #include "iostat.h" #define NFSDBG_FACILITY NFSDBG_FILE @@ -55,6 +56,8 @@ static int nfs_lock(struct file *filp, int cmd, struct file_lock *fl); static int nfs_flock(struct file *filp, int cmd, struct file_lock *fl); static int nfs_setlease(struct file *file, long arg, struct file_lock **fl); +static struct vm_operations_struct nfs_file_vm_ops; + const struct file_operations nfs_file_operations = { .llseek = nfs_file_llseek, .read = do_sync_read, @@ -174,13 +177,38 @@ static loff_t nfs_file_llseek(struct file *filp, loff_t offset, int origin) } /* + * Helper for nfs_file_flush() and nfs_fsync() + * + * Notice that it clears the NFS_CONTEXT_ERROR_WRITE before synching to + * disk, but it retrieves and clears ctx->error after synching, despite + * the two being set at the same time in nfs_context_set_write_error(). + * This is because the former is used to notify the _next_ call to + * nfs_file_write() that a write error occured, and hence cause it to + * fall back to doing a synchronous write. + */ +static int nfs_do_fsync(struct nfs_open_context *ctx, struct inode *inode) +{ + int have_error, status; + int ret = 0; + + have_error = test_and_clear_bit(NFS_CONTEXT_ERROR_WRITE, &ctx->flags); + status = nfs_wb_all(inode); + have_error |= test_bit(NFS_CONTEXT_ERROR_WRITE, &ctx->flags); + if (have_error) + ret = xchg(&ctx->error, 0); + if (!ret) + ret = status; + return ret; +} + +/* * Flush all dirty pages, and check for write errors. * */ static int nfs_file_flush(struct file *file, fl_owner_t id) { - struct nfs_open_context *ctx = (struct nfs_open_context *)file->private_data; + struct nfs_open_context *ctx = nfs_file_open_context(file); struct inode *inode = file->f_path.dentry->d_inode; int status; @@ -189,16 +217,11 @@ nfs_file_flush(struct file *file, fl_owner_t id) if ((file->f_mode & FMODE_WRITE) == 0) return 0; nfs_inc_stats(inode, NFSIOS_VFSFLUSH); - lock_kernel(); + /* Ensure that data+attribute caches are up to date after close() */ - status = nfs_wb_all(inode); - if (!status) { - status = ctx->error; - ctx->error = 0; - if (!status) - nfs_revalidate_inode(NFS_SERVER(inode), inode); - } - unlock_kernel(); + status = nfs_do_fsync(ctx, inode); + if (!status) + nfs_revalidate_inode(NFS_SERVER(inode), inode); return status; } @@ -257,8 +280,11 @@ nfs_file_mmap(struct file * file, struct vm_area_struct * vma) dentry->d_parent->d_name.name, dentry->d_name.name); status = nfs_revalidate_mapping(inode, file->f_mapping); - if (!status) - status = generic_file_mmap(file, vma); + if (!status) { + vma->vm_ops = &nfs_file_vm_ops; + vma->vm_flags |= VM_CAN_NONLINEAR; + file_accessed(file); + } return status; } @@ -270,21 +296,13 @@ nfs_file_mmap(struct file * file, struct vm_area_struct * vma) static int nfs_fsync(struct file *file, struct dentry *dentry, int datasync) { - struct nfs_open_context *ctx = (struct nfs_open_context *)file->private_data; + struct nfs_open_context *ctx = nfs_file_open_context(file); struct inode *inode = dentry->d_inode; - int status; dfprintk(VFS, "nfs: fsync(%s/%ld)\n", inode->i_sb->s_id, inode->i_ino); nfs_inc_stats(inode, NFSIOS_VFSFSYNC); - lock_kernel(); - status = nfs_wb_all(inode); - if (!status) { - status = ctx->error; - ctx->error = 0; - } - unlock_kernel(); - return status; + return nfs_do_fsync(ctx, inode); } /* @@ -333,7 +351,7 @@ static int nfs_launder_page(struct page *page) const struct address_space_operations nfs_file_aops = { .readpage = nfs_readpage, .readpages = nfs_readpages, - .set_page_dirty = nfs_set_page_dirty, + .set_page_dirty = __set_page_dirty_nobuffers, .writepage = nfs_writepage, .writepages = nfs_writepages, .prepare_write = nfs_prepare_write, @@ -346,6 +364,43 @@ const struct address_space_operations nfs_file_aops = { .launder_page = nfs_launder_page, }; +static int nfs_vm_page_mkwrite(struct vm_area_struct *vma, struct page *page) +{ + struct file *filp = vma->vm_file; + unsigned pagelen; + int ret = -EINVAL; + + lock_page(page); + if (page->mapping != vma->vm_file->f_path.dentry->d_inode->i_mapping) + goto out_unlock; + pagelen = nfs_page_length(page); + if (pagelen == 0) + goto out_unlock; + ret = nfs_prepare_write(filp, page, 0, pagelen); + if (!ret) + ret = nfs_commit_write(filp, page, 0, pagelen); +out_unlock: + unlock_page(page); + return ret; +} + +static struct vm_operations_struct nfs_file_vm_ops = { + .fault = filemap_fault, + .page_mkwrite = nfs_vm_page_mkwrite, +}; + +static int nfs_need_sync_write(struct file *filp, struct inode *inode) +{ + struct nfs_open_context *ctx; + + if (IS_SYNC(inode) || (filp->f_flags & O_SYNC)) + return 1; + ctx = nfs_file_open_context(filp); + if (test_bit(NFS_CONTEXT_ERROR_WRITE, &ctx->flags)) + return 1; + return 0; +} + static ssize_t nfs_file_write(struct kiocb *iocb, const struct iovec *iov, unsigned long nr_segs, loff_t pos) { @@ -382,8 +437,8 @@ static ssize_t nfs_file_write(struct kiocb *iocb, const struct iovec *iov, nfs_add_stats(inode, NFSIOS_NORMALWRITTENBYTES, count); result = generic_file_aio_write(iocb, iov, nr_segs, pos); /* Return error values for O_SYNC and IS_SYNC() */ - if (result >= 0 && (IS_SYNC(inode) || (iocb->ki_filp->f_flags & O_SYNC))) { - int err = nfs_fsync(iocb->ki_filp, dentry, 1); + if (result >= 0 && nfs_need_sync_write(iocb->ki_filp, inode)) { + int err = nfs_do_fsync(nfs_file_open_context(iocb->ki_filp), inode); if (err < 0) result = err; } diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c index 71a49c3acab..035c769b715 100644 --- a/fs/nfs/inode.c +++ b/fs/nfs/inode.c @@ -49,6 +49,11 @@ #define NFSDBG_FACILITY NFSDBG_VFS +#define NFS_64_BIT_INODE_NUMBERS_ENABLED 1 + +/* Default is to see 64-bit inode numbers */ +static int enable_ino64 = NFS_64_BIT_INODE_NUMBERS_ENABLED; + static void nfs_invalidate_inode(struct inode *); static int nfs_update_inode(struct inode *, struct nfs_fattr *); @@ -62,6 +67,25 @@ nfs_fattr_to_ino_t(struct nfs_fattr *fattr) return nfs_fileid_to_ino_t(fattr->fileid); } +/** + * nfs_compat_user_ino64 - returns the user-visible inode number + * @fileid: 64-bit fileid + * + * This function returns a 32-bit inode number if the boot parameter + * nfs.enable_ino64 is zero. + */ +u64 nfs_compat_user_ino64(u64 fileid) +{ + int ino; + + if (enable_ino64) + return fileid; + ino = fileid; + if (sizeof(ino) < sizeof(fileid)) + ino ^= fileid >> (sizeof(fileid)-sizeof(ino)) * 8; + return ino; +} + int nfs_write_inode(struct inode *inode, int sync) { int ret; @@ -85,7 +109,6 @@ void nfs_clear_inode(struct inode *inode) */ BUG_ON(nfs_have_writebacks(inode)); BUG_ON(!list_empty(&NFS_I(inode)->open_files)); - BUG_ON(atomic_read(&NFS_I(inode)->data_updates) != 0); nfs_zap_acl_cache(inode); nfs_access_zap_cache(inode); } @@ -118,8 +141,8 @@ static void nfs_zap_caches_locked(struct inode *inode) nfs_inc_stats(inode, NFSIOS_ATTRINVALIDATE); - NFS_ATTRTIMEO(inode) = NFS_MINATTRTIMEO(inode); - NFS_ATTRTIMEO_UPDATE(inode) = jiffies; + nfsi->attrtimeo = NFS_MINATTRTIMEO(inode); + nfsi->attrtimeo_timestamp = jiffies; memset(NFS_COOKIEVERF(inode), 0, sizeof(NFS_COOKIEVERF(inode))); if (S_ISREG(mode) || S_ISDIR(mode) || S_ISLNK(mode)) @@ -156,6 +179,13 @@ static void nfs_zap_acl_cache(struct inode *inode) spin_unlock(&inode->i_lock); } +void nfs_invalidate_atime(struct inode *inode) +{ + spin_lock(&inode->i_lock); + NFS_I(inode)->cache_validity |= NFS_INO_INVALID_ATIME; + spin_unlock(&inode->i_lock); +} + /* * Invalidate, but do not unhash, the inode. * NB: must be called with inode->i_lock held! @@ -338,7 +368,6 @@ nfs_setattr(struct dentry *dentry, struct iattr *attr) return 0; lock_kernel(); - nfs_begin_data_update(inode); /* Write all dirty data */ if (S_ISREG(inode->i_mode)) { filemap_write_and_wait(inode->i_mapping); @@ -352,7 +381,6 @@ nfs_setattr(struct dentry *dentry, struct iattr *attr) error = NFS_PROTO(inode)->setattr(dentry, &fattr, attr); if (error == 0) nfs_refresh_inode(inode, &fattr); - nfs_end_data_update(inode); unlock_kernel(); return error; } @@ -431,7 +459,7 @@ int nfs_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat) /* Flush out writes to the server in order to update c/mtime */ if (S_ISREG(inode->i_mode)) - nfs_sync_mapping_range(inode->i_mapping, 0, 0, FLUSH_NOCOMMIT); + nfs_wb_nocommit(inode); /* * We may force a getattr if the user cares about atime. @@ -450,8 +478,10 @@ int nfs_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat) err = __nfs_revalidate_inode(NFS_SERVER(inode), inode); else err = nfs_revalidate_inode(NFS_SERVER(inode), inode); - if (!err) + if (!err) { generic_fillattr(inode, stat); + stat->ino = nfs_compat_user_ino64(NFS_FILEID(inode)); + } return err; } @@ -536,7 +566,7 @@ struct nfs_open_context *nfs_find_open_context(struct inode *inode, struct rpc_c static void nfs_file_clear_open_context(struct file *filp) { struct inode *inode = filp->f_path.dentry->d_inode; - struct nfs_open_context *ctx = (struct nfs_open_context *)filp->private_data; + struct nfs_open_context *ctx = nfs_file_open_context(filp); if (ctx) { filp->private_data = NULL; @@ -598,16 +628,10 @@ __nfs_revalidate_inode(struct nfs_server *server, struct inode *inode) status = nfs_wait_on_inode(inode); if (status < 0) goto out; - if (NFS_STALE(inode)) { - status = -ESTALE; - /* Do we trust the cached ESTALE? */ - if (NFS_ATTRTIMEO(inode) != 0) { - if (nfsi->cache_validity & (NFS_INO_INVALID_ATTR|NFS_INO_INVALID_ATIME)) { - /* no */ - } else - goto out; - } - } + + status = -ESTALE; + if (NFS_STALE(inode)) + goto out; status = NFS_PROTO(inode)->getattr(server, NFS_FH(inode), &fattr); if (status != 0) { @@ -654,7 +678,7 @@ int nfs_attribute_timeout(struct inode *inode) if (nfs_have_delegation(inode, FMODE_READ)) return 0; - return time_after(jiffies, nfsi->read_cache_jiffies+nfsi->attrtimeo); + return !time_in_range(jiffies, nfsi->read_cache_jiffies, nfsi->read_cache_jiffies + nfsi->attrtimeo); } /** @@ -683,11 +707,8 @@ static int nfs_invalidate_mapping_nolock(struct inode *inode, struct address_spa } spin_lock(&inode->i_lock); nfsi->cache_validity &= ~NFS_INO_INVALID_DATA; - if (S_ISDIR(inode->i_mode)) { + if (S_ISDIR(inode->i_mode)) memset(nfsi->cookieverf, 0, sizeof(nfsi->cookieverf)); - /* This ensures we revalidate child dentries */ - nfsi->cache_change_attribute = jiffies; - } spin_unlock(&inode->i_lock); nfs_inc_stats(inode, NFSIOS_DATAINVALIDATE); dfprintk(PAGECACHE, "NFS: (%s/%Ld) data cache invalidated\n", @@ -756,56 +777,27 @@ out: return ret; } -/** - * nfs_begin_data_update - * @inode - pointer to inode - * Declare that a set of operations will update file data on the server - */ -void nfs_begin_data_update(struct inode *inode) -{ - atomic_inc(&NFS_I(inode)->data_updates); -} - -/** - * nfs_end_data_update - * @inode - pointer to inode - * Declare end of the operations that will update file data - * This will mark the inode as immediately needing revalidation - * of its attribute cache. - */ -void nfs_end_data_update(struct inode *inode) -{ - struct nfs_inode *nfsi = NFS_I(inode); - - /* Directories: invalidate page cache */ - if (S_ISDIR(inode->i_mode)) { - spin_lock(&inode->i_lock); - nfsi->cache_validity |= NFS_INO_INVALID_DATA; - spin_unlock(&inode->i_lock); - } - nfsi->cache_change_attribute = jiffies; - atomic_dec(&nfsi->data_updates); -} - static void nfs_wcc_update_inode(struct inode *inode, struct nfs_fattr *fattr) { struct nfs_inode *nfsi = NFS_I(inode); - unsigned long now = jiffies; + if ((fattr->valid & NFS_ATTR_WCC_V4) != 0 && + nfsi->change_attr == fattr->pre_change_attr) { + nfsi->change_attr = fattr->change_attr; + if (S_ISDIR(inode->i_mode)) + nfsi->cache_validity |= NFS_INO_INVALID_DATA; + } /* If we have atomic WCC data, we may update some attributes */ if ((fattr->valid & NFS_ATTR_WCC) != 0) { - if (timespec_equal(&inode->i_ctime, &fattr->pre_ctime)) { + if (timespec_equal(&inode->i_ctime, &fattr->pre_ctime)) memcpy(&inode->i_ctime, &fattr->ctime, sizeof(inode->i_ctime)); - nfsi->cache_change_attribute = now; - } if (timespec_equal(&inode->i_mtime, &fattr->pre_mtime)) { memcpy(&inode->i_mtime, &fattr->mtime, sizeof(inode->i_mtime)); - nfsi->cache_change_attribute = now; + if (S_ISDIR(inode->i_mode)) + nfsi->cache_validity |= NFS_INO_INVALID_DATA; } - if (inode->i_size == fattr->pre_size && nfsi->npages == 0) { + if (inode->i_size == fattr->pre_size && nfsi->npages == 0) inode->i_size = fattr->size; - nfsi->cache_change_attribute = now; - } } } @@ -822,7 +814,7 @@ static int nfs_check_inode_attributes(struct inode *inode, struct nfs_fattr *fat { struct nfs_inode *nfsi = NFS_I(inode); loff_t cur_size, new_isize; - int data_unstable; + unsigned long invalid = 0; /* Has the inode gone and changed behind our back? */ @@ -831,37 +823,41 @@ static int nfs_check_inode_attributes(struct inode *inode, struct nfs_fattr *fat return -EIO; } - /* Are we in the process of updating data on the server? */ - data_unstable = nfs_caches_unstable(inode); - /* Do atomic weak cache consistency updates */ nfs_wcc_update_inode(inode, fattr); if ((fattr->valid & NFS_ATTR_FATTR_V4) != 0 && nfsi->change_attr != fattr->change_attr) - nfsi->cache_validity |= NFS_INO_INVALID_ATTR|NFS_INO_REVAL_PAGECACHE; + invalid |= NFS_INO_INVALID_ATTR|NFS_INO_REVAL_PAGECACHE; /* Verify a few of the more important attributes */ if (!timespec_equal(&inode->i_mtime, &fattr->mtime)) - nfsi->cache_validity |= NFS_INO_INVALID_ATTR|NFS_INO_REVAL_PAGECACHE; + invalid |= NFS_INO_INVALID_ATTR|NFS_INO_REVAL_PAGECACHE; cur_size = i_size_read(inode); new_isize = nfs_size_to_loff_t(fattr->size); if (cur_size != new_isize && nfsi->npages == 0) - nfsi->cache_validity |= NFS_INO_INVALID_ATTR|NFS_INO_REVAL_PAGECACHE; + invalid |= NFS_INO_INVALID_ATTR|NFS_INO_REVAL_PAGECACHE; /* Have any file permissions changed? */ if ((inode->i_mode & S_IALLUGO) != (fattr->mode & S_IALLUGO) || inode->i_uid != fattr->uid || inode->i_gid != fattr->gid) - nfsi->cache_validity |= NFS_INO_INVALID_ATTR | NFS_INO_INVALID_ACCESS | NFS_INO_INVALID_ACL; + invalid |= NFS_INO_INVALID_ATTR | NFS_INO_INVALID_ACCESS | NFS_INO_INVALID_ACL; /* Has the link count changed? */ if (inode->i_nlink != fattr->nlink) - nfsi->cache_validity |= NFS_INO_INVALID_ATTR; + invalid |= NFS_INO_INVALID_ATTR; if (!timespec_equal(&inode->i_atime, &fattr->atime)) - nfsi->cache_validity |= NFS_INO_INVALID_ATIME; + invalid |= NFS_INO_INVALID_ATIME; + + if (invalid != 0) + nfsi->cache_validity |= invalid; + else + nfsi->cache_validity &= ~(NFS_INO_INVALID_ATTR + | NFS_INO_INVALID_ATIME + | NFS_INO_REVAL_PAGECACHE); nfsi->read_cache_jiffies = fattr->time_start; return 0; @@ -911,17 +907,41 @@ int nfs_refresh_inode(struct inode *inode, struct nfs_fattr *fattr) int nfs_post_op_update_inode(struct inode *inode, struct nfs_fattr *fattr) { struct nfs_inode *nfsi = NFS_I(inode); - int status = 0; spin_lock(&inode->i_lock); - if (unlikely((fattr->valid & NFS_ATTR_FATTR) == 0)) { - nfsi->cache_validity |= NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ATTR|NFS_INO_REVAL_PAGECACHE; - goto out; - } - status = nfs_update_inode(inode, fattr); -out: + nfsi->cache_validity |= NFS_INO_INVALID_ATTR|NFS_INO_REVAL_PAGECACHE; + if (S_ISDIR(inode->i_mode)) + nfsi->cache_validity |= NFS_INO_INVALID_DATA; spin_unlock(&inode->i_lock); - return status; + return nfs_refresh_inode(inode, fattr); +} + +/** + * nfs_post_op_update_inode_force_wcc - try to update the inode attribute cache + * @inode - pointer to inode + * @fattr - updated attributes + * + * After an operation that has changed the inode metadata, mark the + * attribute cache as being invalid, then try to update it. Fake up + * weak cache consistency data, if none exist. + * + * This function is mainly designed to be used by the ->write_done() functions. + */ +int nfs_post_op_update_inode_force_wcc(struct inode *inode, struct nfs_fattr *fattr) +{ + if ((fattr->valid & NFS_ATTR_FATTR_V4) != 0 && + (fattr->valid & NFS_ATTR_WCC_V4) == 0) { + fattr->pre_change_attr = NFS_I(inode)->change_attr; + fattr->valid |= NFS_ATTR_WCC_V4; + } + if ((fattr->valid & NFS_ATTR_FATTR) != 0 && + (fattr->valid & NFS_ATTR_WCC) == 0) { + memcpy(&fattr->pre_ctime, &inode->i_ctime, sizeof(fattr->pre_ctime)); + memcpy(&fattr->pre_mtime, &inode->i_mtime, sizeof(fattr->pre_mtime)); + fattr->pre_size = inode->i_size; + fattr->valid |= NFS_ATTR_WCC; + } + return nfs_post_op_update_inode(inode, fattr); } /* @@ -941,9 +961,8 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr) struct nfs_server *server; struct nfs_inode *nfsi = NFS_I(inode); loff_t cur_isize, new_isize; - unsigned int invalid = 0; + unsigned long invalid = 0; unsigned long now = jiffies; - int data_stable; dfprintk(VFS, "NFS: %s(%s/%ld ct=%d info=0x%x)\n", __FUNCTION__, inode->i_sb->s_id, inode->i_ino, @@ -968,57 +987,51 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr) * Update the read time so we don't revalidate too often. */ nfsi->read_cache_jiffies = fattr->time_start; - nfsi->last_updated = now; - /* Fix a wraparound issue with nfsi->cache_change_attribute */ - if (time_before(now, nfsi->cache_change_attribute)) - nfsi->cache_change_attribute = now - 600*HZ; - - /* Are we racing with known updates of the metadata on the server? */ - data_stable = nfs_verify_change_attribute(inode, fattr->time_start); - if (data_stable) - nfsi->cache_validity &= ~(NFS_INO_INVALID_ATTR|NFS_INO_REVAL_PAGECACHE|NFS_INO_INVALID_ATIME); + nfsi->cache_validity &= ~(NFS_INO_INVALID_ATTR | NFS_INO_INVALID_ATIME + | NFS_INO_REVAL_PAGECACHE); /* Do atomic weak cache consistency updates */ nfs_wcc_update_inode(inode, fattr); + /* More cache consistency checks */ + if (!(fattr->valid & NFS_ATTR_FATTR_V4)) { + /* NFSv2/v3: Check if the mtime agrees */ + if (!timespec_equal(&inode->i_mtime, &fattr->mtime)) { + dprintk("NFS: mtime change on server for file %s/%ld\n", + inode->i_sb->s_id, inode->i_ino); + invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_DATA; + nfsi->cache_change_attribute = now; + } + /* If ctime has changed we should definitely clear access+acl caches */ + if (!timespec_equal(&inode->i_ctime, &fattr->ctime)) + invalid |= NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ACL; + } else if (nfsi->change_attr != fattr->change_attr) { + dprintk("NFS: change_attr change on server for file %s/%ld\n", + inode->i_sb->s_id, inode->i_ino); + invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_DATA|NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ACL; + nfsi->cache_change_attribute = now; + } + /* Check if our cached file size is stale */ new_isize = nfs_size_to_loff_t(fattr->size); cur_isize = i_size_read(inode); if (new_isize != cur_isize) { - /* Do we perhaps have any outstanding writes? */ - if (nfsi->npages == 0) { - /* No, but did we race with nfs_end_data_update()? */ - if (data_stable) { - inode->i_size = new_isize; - invalid |= NFS_INO_INVALID_DATA; - } - invalid |= NFS_INO_INVALID_ATTR; - } else if (new_isize > cur_isize) { + /* Do we perhaps have any outstanding writes, or has + * the file grown beyond our last write? */ + if (nfsi->npages == 0 || new_isize > cur_isize) { inode->i_size = new_isize; invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_DATA; } - nfsi->cache_change_attribute = now; dprintk("NFS: isize change on server for file %s/%ld\n", inode->i_sb->s_id, inode->i_ino); } - /* Check if the mtime agrees */ - if (!timespec_equal(&inode->i_mtime, &fattr->mtime)) { - memcpy(&inode->i_mtime, &fattr->mtime, sizeof(inode->i_mtime)); - dprintk("NFS: mtime change on server for file %s/%ld\n", - inode->i_sb->s_id, inode->i_ino); - invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_DATA; - nfsi->cache_change_attribute = now; - } - /* If ctime has changed we should definitely clear access+acl caches */ - if (!timespec_equal(&inode->i_ctime, &fattr->ctime)) { - invalid |= NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ACL; - memcpy(&inode->i_ctime, &fattr->ctime, sizeof(inode->i_ctime)); - nfsi->cache_change_attribute = now; - } + memcpy(&inode->i_mtime, &fattr->mtime, sizeof(inode->i_mtime)); + memcpy(&inode->i_ctime, &fattr->ctime, sizeof(inode->i_ctime)); memcpy(&inode->i_atime, &fattr->atime, sizeof(inode->i_atime)); + nfsi->change_attr = fattr->change_attr; if ((inode->i_mode & S_IALLUGO) != (fattr->mode & S_IALLUGO) || inode->i_uid != fattr->uid || @@ -1039,31 +1052,29 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr) inode->i_blocks = fattr->du.nfs2.blocks; } - if ((fattr->valid & NFS_ATTR_FATTR_V4) != 0 && - nfsi->change_attr != fattr->change_attr) { - dprintk("NFS: change_attr change on server for file %s/%ld\n", - inode->i_sb->s_id, inode->i_ino); - nfsi->change_attr = fattr->change_attr; - invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_DATA|NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ACL; - nfsi->cache_change_attribute = now; - } - /* Update attrtimeo value if we're out of the unstable period */ if (invalid & NFS_INO_INVALID_ATTR) { nfs_inc_stats(inode, NFSIOS_ATTRINVALIDATE); nfsi->attrtimeo = NFS_MINATTRTIMEO(inode); nfsi->attrtimeo_timestamp = now; - } else if (time_after(now, nfsi->attrtimeo_timestamp+nfsi->attrtimeo)) { - if ((nfsi->attrtimeo <<= 1) > NFS_MAXATTRTIMEO(inode)) - nfsi->attrtimeo = NFS_MAXATTRTIMEO(inode); - nfsi->attrtimeo_timestamp = now; + nfsi->last_updated = now; + } else { + if (!time_in_range(now, nfsi->attrtimeo_timestamp, nfsi->attrtimeo_timestamp + nfsi->attrtimeo)) { + if ((nfsi->attrtimeo <<= 1) > NFS_MAXATTRTIMEO(inode)) + nfsi->attrtimeo = NFS_MAXATTRTIMEO(inode); + nfsi->attrtimeo_timestamp = now; + } + /* + * Avoid jiffy wraparound issues with nfsi->last_updated + */ + if (!time_in_range(nfsi->last_updated, nfsi->read_cache_jiffies, now)) + nfsi->last_updated = nfsi->read_cache_jiffies; } + invalid &= ~NFS_INO_INVALID_ATTR; /* Don't invalidate the data if we were to blame */ if (!(S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode))) 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) || (nfsi->cache_validity & NFS_INO_REVAL_FORCED)) nfsi->cache_validity |= invalid; @@ -1152,7 +1163,6 @@ static void init_once(void * foo, struct kmem_cache * cachep, unsigned long flag 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->ncommit = 0; nfsi->npages = 0; nfs4_init_once(nfsi); @@ -1249,6 +1259,7 @@ static void __exit exit_nfs_fs(void) /* Not quite true; I just maintain it */ MODULE_AUTHOR("Olaf Kirch <okir@monad.swb.de>"); MODULE_LICENSE("GPL"); +module_param(enable_ino64, bool, 0644); module_init(init_nfs_fs) module_exit(exit_nfs_fs) diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h index 76cf55d5710..f3acf48412b 100644 --- a/fs/nfs/internal.h +++ b/fs/nfs/internal.h @@ -5,8 +5,6 @@ #include <linux/mount.h> struct nfs_string; -struct nfs_mount_data; -struct nfs4_mount_data; /* Maximum number of readahead requests * FIXME: this should really be a sysctl so that users may tune it to suit @@ -27,20 +25,50 @@ struct nfs_clone_mount { rpc_authflavor_t authflavor; }; +/* + * In-kernel mount arguments + */ +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; + char *hostname; + 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; +}; + /* client.c */ extern struct rpc_program nfs_program; extern void nfs_put_client(struct nfs_client *); extern struct nfs_client *nfs_find_client(const struct sockaddr_in *, int); -extern struct nfs_server *nfs_create_server(const struct nfs_mount_data *, - struct nfs_fh *); -extern struct nfs_server *nfs4_create_server(const struct nfs4_mount_data *, - const char *, - const struct sockaddr_in *, - const char *, - const char *, - rpc_authflavor_t, - struct nfs_fh *); +extern struct nfs_server *nfs_create_server( + const struct nfs_parsed_mount_data *, + struct nfs_fh *); +extern struct nfs_server *nfs4_create_server( + const struct nfs_parsed_mount_data *, + struct nfs_fh *); extern struct nfs_server *nfs4_create_referral_server(struct nfs_clone_mount *, struct nfs_fh *); extern void nfs_free_server(struct nfs_server *server); diff --git a/fs/nfs/nfs2xdr.c b/fs/nfs/nfs2xdr.c index c5fce756720..668ab96c7b5 100644 --- a/fs/nfs/nfs2xdr.c +++ b/fs/nfs/nfs2xdr.c @@ -251,6 +251,7 @@ nfs_xdr_readargs(struct rpc_rqst *req, __be32 *p, struct nfs_readargs *args) replen = (RPC_REPHDRSIZE + auth->au_rslack + NFS_readres_sz) << 2; xdr_inline_pages(&req->rq_rcv_buf, replen, args->pages, args->pgbase, count); + req->rq_rcv_buf.flags |= XDRBUF_READ; return 0; } @@ -271,7 +272,7 @@ nfs_xdr_readres(struct rpc_rqst *req, __be32 *p, struct nfs_readres *res) res->eof = 0; hdrlen = (u8 *) p - (u8 *) iov->iov_base; if (iov->iov_len < hdrlen) { - printk(KERN_WARNING "NFS: READ reply header overflowed:" + dprintk("NFS: READ reply header overflowed:" "length %d > %Zu\n", hdrlen, iov->iov_len); return -errno_NFSERR_IO; } else if (iov->iov_len != hdrlen) { @@ -281,7 +282,7 @@ nfs_xdr_readres(struct rpc_rqst *req, __be32 *p, struct nfs_readres *res) recvd = req->rq_rcv_buf.len - hdrlen; if (count > recvd) { - printk(KERN_WARNING "NFS: server cheating in read reply: " + dprintk("NFS: server cheating in read reply: " "count %d > recvd %d\n", count, recvd); count = recvd; } @@ -313,6 +314,7 @@ nfs_xdr_writeargs(struct rpc_rqst *req, __be32 *p, struct nfs_writeargs *args) /* Copy the page array */ xdr_encode_pages(sndbuf, args->pages, args->pgbase, count); + sndbuf->flags |= XDRBUF_WRITE; return 0; } @@ -431,7 +433,7 @@ nfs_xdr_readdirres(struct rpc_rqst *req, __be32 *p, void *dummy) hdrlen = (u8 *) p - (u8 *) iov->iov_base; if (iov->iov_len < hdrlen) { - printk(KERN_WARNING "NFS: READDIR reply header overflowed:" + dprintk("NFS: READDIR reply header overflowed:" "length %d > %Zu\n", hdrlen, iov->iov_len); return -errno_NFSERR_IO; } else if (iov->iov_len != hdrlen) { @@ -454,7 +456,7 @@ nfs_xdr_readdirres(struct rpc_rqst *req, __be32 *p, void *dummy) len = ntohl(*p++); p += XDR_QUADLEN(len) + 1; /* name plus cookie */ if (len > NFS2_MAXNAMLEN) { - printk(KERN_WARNING "NFS: giant filename in readdir (len 0x%x)!\n", + dprintk("NFS: giant filename in readdir (len 0x%x)!\n", len); goto err_unmap; } @@ -471,7 +473,7 @@ nfs_xdr_readdirres(struct rpc_rqst *req, __be32 *p, void *dummy) entry[0] = entry[1] = 0; /* truncate listing ? */ if (!nr) { - printk(KERN_NOTICE "NFS: readdir reply truncated!\n"); + dprintk("NFS: readdir reply truncated!\n"); entry[1] = 1; } goto out; @@ -583,12 +585,12 @@ nfs_xdr_readlinkres(struct rpc_rqst *req, __be32 *p, void *dummy) /* Convert length of symlink */ len = ntohl(*p++); if (len >= rcvbuf->page_len || len <= 0) { - dprintk(KERN_WARNING "nfs: server returned giant symlink!\n"); + dprintk("nfs: server returned giant symlink!\n"); return -ENAMETOOLONG; } hdrlen = (u8 *) p - (u8 *) iov->iov_base; if (iov->iov_len < hdrlen) { - printk(KERN_WARNING "NFS: READLINK reply header overflowed:" + dprintk("NFS: READLINK reply header overflowed:" "length %d > %Zu\n", hdrlen, iov->iov_len); return -errno_NFSERR_IO; } else if (iov->iov_len != hdrlen) { @@ -597,7 +599,7 @@ nfs_xdr_readlinkres(struct rpc_rqst *req, __be32 *p, void *dummy) } recvd = req->rq_rcv_buf.len - hdrlen; if (recvd < len) { - printk(KERN_WARNING "NFS: server cheating in readlink reply: " + dprintk("NFS: server cheating in readlink reply: " "count %u > recvd %u\n", len, recvd); return -EIO; } @@ -695,7 +697,7 @@ nfs_stat_to_errno(int stat) if (nfs_errtbl[i].stat == stat) return nfs_errtbl[i].errno; } - printk(KERN_ERR "nfs_stat_to_errno: bad nfs status return value: %d\n", stat); + dprintk("nfs_stat_to_errno: bad nfs status return value: %d\n", stat); return nfs_errtbl[i].errno; } diff --git a/fs/nfs/nfs3acl.c b/fs/nfs/nfs3acl.c index 7322da4d205..9b7362565c0 100644 --- a/fs/nfs/nfs3acl.c +++ b/fs/nfs/nfs3acl.c @@ -317,13 +317,11 @@ static int nfs3_proc_setacls(struct inode *inode, struct posix_acl *acl, } dprintk("NFS call setacl\n"); - nfs_begin_data_update(inode); msg.rpc_proc = &server->client_acl->cl_procinfo[ACLPROC3_SETACL]; status = rpc_call_sync(server->client_acl, &msg, 0); spin_lock(&inode->i_lock); NFS_I(inode)->cache_validity |= NFS_INO_INVALID_ACCESS; spin_unlock(&inode->i_lock); - nfs_end_data_update(inode); dprintk("NFS reply setacl: %d\n", status); /* pages may have been allocated at the xdr layer. */ diff --git a/fs/nfs/nfs3proc.c b/fs/nfs/nfs3proc.c index c7ca5d70870..4cdc2361a66 100644 --- a/fs/nfs/nfs3proc.c +++ b/fs/nfs/nfs3proc.c @@ -166,6 +166,7 @@ nfs3_proc_lookup(struct inode *dir, struct qstr *name, nfs_fattr_init(&dir_attr); nfs_fattr_init(fattr); status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); + nfs_refresh_inode(dir, &dir_attr); if (status >= 0 && !(fattr->valid & NFS_ATTR_FATTR)) { msg.rpc_proc = &nfs3_procedures[NFS3PROC_GETATTR]; msg.rpc_argp = fhandle; @@ -173,8 +174,6 @@ nfs3_proc_lookup(struct inode *dir, struct qstr *name, status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); } dprintk("NFS reply lookup: %d\n", status); - if (status >= 0) - status = nfs_refresh_inode(dir, &dir_attr); return status; } @@ -607,6 +606,9 @@ nfs3_proc_readdir(struct dentry *dentry, struct rpc_cred *cred, nfs_fattr_init(&dir_attr); status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); + + nfs_invalidate_atime(dir); + nfs_refresh_inode(dir, &dir_attr); dprintk("NFS reply readdir: %d\n", status); return status; @@ -724,9 +726,9 @@ static int nfs3_read_done(struct rpc_task *task, struct nfs_read_data *data) { if (nfs3_async_handle_jukebox(task, data->inode)) return -EAGAIN; - /* Call back common NFS readpage processing */ - if (task->tk_status >= 0) - nfs_refresh_inode(data->inode, &data->fattr); + + nfs_invalidate_atime(data->inode); + nfs_refresh_inode(data->inode, &data->fattr); return 0; } @@ -747,7 +749,7 @@ static int nfs3_write_done(struct rpc_task *task, struct nfs_write_data *data) if (nfs3_async_handle_jukebox(task, data->inode)) return -EAGAIN; if (task->tk_status >= 0) - nfs_post_op_update_inode(data->inode, data->res.fattr); + nfs_post_op_update_inode_force_wcc(data->inode, data->res.fattr); return 0; } @@ -775,8 +777,7 @@ static int nfs3_commit_done(struct rpc_task *task, struct nfs_write_data *data) { if (nfs3_async_handle_jukebox(task, data->inode)) return -EAGAIN; - if (task->tk_status >= 0) - nfs_post_op_update_inode(data->inode, data->res.fattr); + nfs_refresh_inode(data->inode, data->res.fattr); return 0; } diff --git a/fs/nfs/nfs3xdr.c b/fs/nfs/nfs3xdr.c index d9e08f0cf2a..616d3267b7e 100644 --- a/fs/nfs/nfs3xdr.c +++ b/fs/nfs/nfs3xdr.c @@ -346,6 +346,7 @@ nfs3_xdr_readargs(struct rpc_rqst *req, __be32 *p, struct nfs_readargs *args) replen = (RPC_REPHDRSIZE + auth->au_rslack + NFS3_readres_sz) << 2; xdr_inline_pages(&req->rq_rcv_buf, replen, args->pages, args->pgbase, count); + req->rq_rcv_buf.flags |= XDRBUF_READ; return 0; } @@ -367,6 +368,7 @@ nfs3_xdr_writeargs(struct rpc_rqst *req, __be32 *p, struct nfs_writeargs *args) /* Copy the page array */ xdr_encode_pages(sndbuf, args->pages, args->pgbase, count); + sndbuf->flags |= XDRBUF_WRITE; return 0; } @@ -524,7 +526,7 @@ nfs3_xdr_readdirres(struct rpc_rqst *req, __be32 *p, struct nfs3_readdirres *res hdrlen = (u8 *) p - (u8 *) iov->iov_base; if (iov->iov_len < hdrlen) { - printk(KERN_WARNING "NFS: READDIR reply header overflowed:" + dprintk("NFS: READDIR reply header overflowed:" "length %d > %Zu\n", hdrlen, iov->iov_len); return -errno_NFSERR_IO; } else if (iov->iov_len != hdrlen) { @@ -547,7 +549,7 @@ nfs3_xdr_readdirres(struct rpc_rqst *req, __be32 *p, struct nfs3_readdirres *res len = ntohl(*p++); /* string length */ p += XDR_QUADLEN(len) + 2; /* name + cookie */ if (len > NFS3_MAXNAMLEN) { - printk(KERN_WARNING "NFS: giant filename in readdir (len %x)!\n", + dprintk("NFS: giant filename in readdir (len %x)!\n", len); goto err_unmap; } @@ -567,7 +569,7 @@ nfs3_xdr_readdirres(struct rpc_rqst *req, __be32 *p, struct nfs3_readdirres *res goto short_pkt; len = ntohl(*p++); if (len > NFS3_FHSIZE) { - printk(KERN_WARNING "NFS: giant filehandle in " + dprintk("NFS: giant filehandle in " "readdir (len %x)!\n", len); goto err_unmap; } @@ -588,7 +590,7 @@ nfs3_xdr_readdirres(struct rpc_rqst *req, __be32 *p, struct nfs3_readdirres *res entry[0] = entry[1] = 0; /* truncate listing ? */ if (!nr) { - printk(KERN_NOTICE "NFS: readdir reply truncated!\n"); + dprintk("NFS: readdir reply truncated!\n"); entry[1] = 1; } goto out; @@ -826,22 +828,23 @@ nfs3_xdr_readlinkres(struct rpc_rqst *req, __be32 *p, struct nfs_fattr *fattr) /* Convert length of symlink */ len = ntohl(*p++); if (len >= rcvbuf->page_len || len <= 0) { - dprintk(KERN_WARNING "nfs: server returned giant symlink!\n"); + dprintk("nfs: server returned giant symlink!\n"); return -ENAMETOOLONG; } hdrlen = (u8 *) p - (u8 *) iov->iov_base; if (iov->iov_len < hdrlen) { - printk(KERN_WARNING "NFS: READLINK reply header overflowed:" + dprintk("NFS: READLINK reply header overflowed:" "length %d > %Zu\n", hdrlen, iov->iov_len); return -errno_NFSERR_IO; } else if (iov->iov_len != hdrlen) { - dprintk("NFS: READLINK header is short. iovec will be shifted.\n"); + dprintk("NFS: READLINK header is short. " + "iovec will be shifted.\n"); xdr_shift_buf(rcvbuf, iov->iov_len - hdrlen); } recvd = req->rq_rcv_buf.len - hdrlen; if (recvd < len) { - printk(KERN_WARNING "NFS: server cheating in readlink reply: " + dprintk("NFS: server cheating in readlink reply: " "count %u > recvd %u\n", len, recvd); return -EIO; } @@ -876,13 +879,13 @@ nfs3_xdr_readres(struct rpc_rqst *req, __be32 *p, struct nfs_readres *res) ocount = ntohl(*p++); if (ocount != count) { - printk(KERN_WARNING "NFS: READ count doesn't match RPC opaque count.\n"); + dprintk("NFS: READ count doesn't match RPC opaque count.\n"); return -errno_NFSERR_IO; } hdrlen = (u8 *) p - (u8 *) iov->iov_base; if (iov->iov_len < hdrlen) { - printk(KERN_WARNING "NFS: READ reply header overflowed:" + dprintk("NFS: READ reply header overflowed:" "length %d > %Zu\n", hdrlen, iov->iov_len); return -errno_NFSERR_IO; } else if (iov->iov_len != hdrlen) { @@ -892,7 +895,7 @@ nfs3_xdr_readres(struct rpc_rqst *req, __be32 *p, struct nfs_readres *res) recvd = req->rq_rcv_buf.len - hdrlen; if (count > recvd) { - printk(KERN_WARNING "NFS: server cheating in read reply: " + dprintk("NFS: server cheating in read reply: " "count %d > recvd %d\n", count, recvd); count = recvd; res->eof = 0; diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 4b90e17555a..cb99fd90a9a 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -62,10 +62,8 @@ struct nfs4_opendata; static int _nfs4_proc_open(struct nfs4_opendata *data); static int nfs4_do_fsinfo(struct nfs_server *, struct nfs_fh *, struct nfs_fsinfo *); 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); static int _nfs4_proc_lookup(struct inode *dir, const struct qstr *name, struct nfs_fh *fhandle, struct nfs_fattr *fattr); static int _nfs4_proc_getattr(struct nfs_server *server, struct nfs_fh *fhandle, struct nfs_fattr *fattr); @@ -177,7 +175,7 @@ static void nfs4_setup_readdir(u64 cookie, __be32 *verifier, struct dentry *dent *p++ = xdr_one; /* bitmap length */ *p++ = htonl(FATTR4_WORD0_FILEID); /* bitmap */ *p++ = htonl(8); /* attribute buffer length */ - p = xdr_encode_hyper(p, dentry->d_inode->i_ino); + p = xdr_encode_hyper(p, NFS_FILEID(dentry->d_inode)); } *p++ = xdr_one; /* next */ @@ -189,7 +187,7 @@ static void nfs4_setup_readdir(u64 cookie, __be32 *verifier, struct dentry *dent *p++ = xdr_one; /* bitmap length */ *p++ = htonl(FATTR4_WORD0_FILEID); /* bitmap */ *p++ = htonl(8); /* attribute buffer length */ - p = xdr_encode_hyper(p, dentry->d_parent->d_inode->i_ino); + p = xdr_encode_hyper(p, NFS_FILEID(dentry->d_parent->d_inode)); readdir->pgbase = (char *)p - (char *)start; readdir->count -= readdir->pgbase; @@ -211,8 +209,9 @@ static void update_changeattr(struct inode *dir, struct nfs4_change_info *cinfo) spin_lock(&dir->i_lock); nfsi->cache_validity |= NFS_INO_INVALID_ATTR|NFS_INO_REVAL_PAGECACHE|NFS_INO_INVALID_DATA; - if (cinfo->before == nfsi->change_attr && cinfo->atomic) - nfsi->change_attr = cinfo->after; + if (!cinfo->atomic || cinfo->before != nfsi->change_attr) + nfsi->cache_change_attribute = jiffies; + nfsi->change_attr = cinfo->after; spin_unlock(&dir->i_lock); } @@ -454,7 +453,7 @@ static struct nfs4_state *nfs4_try_open_cached(struct nfs4_opendata *opendata) 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); + ret = nfs_may_open(state->inode, state->owner->so_cred, open_mode); unlock_kernel(); if (ret != 0) goto out; @@ -948,36 +947,6 @@ static int _nfs4_proc_open(struct nfs4_opendata *data) return 0; } -static int _nfs4_do_access(struct inode *inode, struct rpc_cred *cred, int openflags) -{ - struct nfs_access_entry cache; - int mask = 0; - int status; - - if (openflags & FMODE_READ) - 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; - - /* Be clever: ask server to check for all possible rights */ - cache.mask = MAY_EXEC | MAY_WRITE | MAY_READ; - cache.cred = cred; - cache.jiffies = jiffies; - status = _nfs4_proc_access(inode, &cache); - if (status != 0) - return status; - nfs_access_add_cache(inode, &cache); -out: - if ((cache.mask & mask) == mask) - return 0; - return -EACCES; -} - static int nfs4_recover_expired_lease(struct nfs_server *server) { struct nfs_client *clp = server->nfs_client; @@ -1381,7 +1350,7 @@ static int nfs4_intent_set_file(struct nameidata *nd, struct path *path, struct /* 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, + ret = nfs_may_open(state->inode, state->owner->so_cred, nd->intent.open.flags); if (ret < 0) @@ -1390,7 +1359,7 @@ static int nfs4_intent_set_file(struct nameidata *nd, struct path *path, struct 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 = nfs_file_open_context(filp); ctx->state = state; return 0; } @@ -1428,13 +1397,16 @@ nfs4_atomic_open(struct inode *dir, struct dentry *dentry, struct nameidata *nd) state = nfs4_do_open(dir, &path, nd->intent.open.flags, &attr, cred); put_rpccred(cred); if (IS_ERR(state)) { - if (PTR_ERR(state) == -ENOENT) + if (PTR_ERR(state) == -ENOENT) { d_add(dentry, NULL); + nfs_set_verifier(dentry, nfs_save_change_attribute(dir)); + } return (struct dentry *)state; } res = d_add_unique(dentry, igrab(state->inode)); if (res != NULL) path.dentry = res; + nfs_set_verifier(path.dentry, nfs_save_change_attribute(dir)); nfs4_intent_set_file(nd, &path, state); return res; } @@ -1468,6 +1440,7 @@ nfs4_open_revalidate(struct inode *dir, struct dentry *dentry, int openflags, st } } if (state->inode == dentry->d_inode) { + nfs_set_verifier(dentry, nfs_save_change_attribute(dir)); nfs4_intent_set_file(nd, &path, state); return 1; } @@ -1757,10 +1730,16 @@ static int nfs4_proc_lookup(struct inode *dir, struct qstr *name, struct nfs_fh static int _nfs4_proc_access(struct inode *inode, struct nfs_access_entry *entry) { + struct nfs_server *server = NFS_SERVER(inode); + struct nfs_fattr fattr; struct nfs4_accessargs args = { .fh = NFS_FH(inode), + .bitmask = server->attr_bitmask, + }; + struct nfs4_accessres res = { + .server = server, + .fattr = &fattr, }; - struct nfs4_accessres res = { 0 }; struct rpc_message msg = { .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_ACCESS], .rpc_argp = &args, @@ -1786,6 +1765,7 @@ static int _nfs4_proc_access(struct inode *inode, struct nfs_access_entry *entry if (mode & MAY_EXEC) args.access |= NFS4_ACCESS_EXECUTE; } + nfs_fattr_init(&fattr); status = rpc_call_sync(NFS_CLIENT(inode), &msg, 0); if (!status) { entry->mask = 0; @@ -1795,6 +1775,7 @@ static int _nfs4_proc_access(struct inode *inode, struct nfs_access_entry *entry entry->mask |= MAY_WRITE; if (res.access & (NFS4_ACCESS_LOOKUP|NFS4_ACCESS_EXECUTE)) entry->mask |= MAY_EXEC; + nfs_refresh_inode(inode, &fattr); } return status; } @@ -1900,11 +1881,13 @@ nfs4_proc_create(struct inode *dir, struct dentry *dentry, struct iattr *sattr, } state = nfs4_do_open(dir, &path, flags, sattr, cred); put_rpccred(cred); + d_drop(dentry); if (IS_ERR(state)) { status = PTR_ERR(state); goto out; } - d_instantiate(dentry, igrab(state->inode)); + d_add(dentry, igrab(state->inode)); + nfs_set_verifier(dentry, nfs_save_change_attribute(dir)); if (flags & O_EXCL) { struct nfs_fattr fattr; status = nfs4_do_setattr(state->inode, &fattr, sattr, state); @@ -2218,6 +2201,9 @@ static int _nfs4_proc_readdir(struct dentry *dentry, struct rpc_cred *cred, status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); if (status == 0) memcpy(NFS_COOKIEVERF(dir), res.verifier.data, NFS4_VERIFIER_SIZE); + + nfs_invalidate_atime(dir); + dprintk("%s: returns %d\n", __FUNCTION__, status); return status; } @@ -2414,6 +2400,8 @@ static int nfs4_read_done(struct rpc_task *task, struct nfs_read_data *data) rpc_restart_call(task); return -EAGAIN; } + + nfs_invalidate_atime(data->inode); if (task->tk_status > 0) renew_lease(server, data->timestamp); return 0; @@ -2443,7 +2431,7 @@ static int nfs4_write_done(struct rpc_task *task, struct nfs_write_data *data) } if (task->tk_status >= 0) { renew_lease(NFS_SERVER(inode), data->timestamp); - nfs_post_op_update_inode(inode, data->res.fattr); + nfs_post_op_update_inode_force_wcc(inode, data->res.fattr); } return 0; } @@ -2485,8 +2473,7 @@ static int nfs4_commit_done(struct rpc_task *task, struct nfs_write_data *data) rpc_restart_call(task); return -EAGAIN; } - if (task->tk_status >= 0) - nfs_post_op_update_inode(inode, data->res.fattr); + nfs_refresh_inode(inode, data->res.fattr); return 0; } @@ -3056,7 +3043,7 @@ static int _nfs4_proc_delegreturn(struct inode *inode, struct rpc_cred *cred, co if (status == 0) { status = data->rpc_status; if (status == 0) - nfs_post_op_update_inode(inode, &data->fattr); + nfs_refresh_inode(inode, &data->fattr); } rpc_put_task(task); return status; @@ -3303,7 +3290,7 @@ static int nfs4_proc_unlck(struct nfs4_state *state, int cmd, struct file_lock * status = -ENOMEM; if (seqid == NULL) goto out; - task = nfs4_do_unlck(request, request->fl_file->private_data, lsp, seqid); + task = nfs4_do_unlck(request, nfs_file_open_context(request->fl_file), lsp, seqid); status = PTR_ERR(task); if (IS_ERR(task)) goto out; @@ -3447,7 +3434,7 @@ static int _nfs4_do_setlk(struct nfs4_state *state, int cmd, struct file_lock *f int ret; dprintk("%s: begin!\n", __FUNCTION__); - data = nfs4_alloc_lockdata(fl, fl->fl_file->private_data, + data = nfs4_alloc_lockdata(fl, nfs_file_open_context(fl->fl_file), fl->fl_u.nfs4_fl.owner); if (data == NULL) return -ENOMEM; @@ -3573,7 +3560,7 @@ nfs4_proc_lock(struct file *filp, int cmd, struct file_lock *request) int status; /* verify open state */ - ctx = (struct nfs_open_context *)filp->private_data; + ctx = nfs_file_open_context(filp); state = ctx->state; if (request->fl_start < 0 || request->fl_end < 0) diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c index 3e4adf8c831..bfb36261cec 100644 --- a/fs/nfs/nfs4state.c +++ b/fs/nfs/nfs4state.c @@ -774,7 +774,7 @@ static int nfs4_reclaim_locks(struct nfs4_state_recovery_ops *ops, struct nfs4_s for (fl = inode->i_flock; fl != 0; fl = fl->fl_next) { if (!(fl->fl_flags & (FL_POSIX|FL_FLOCK))) continue; - if (((struct nfs_open_context *)fl->fl_file->private_data)->state != state) + if (nfs_file_open_context(fl->fl_file)->state != state) continue; status = ops->recover_lock(state, fl); if (status >= 0) diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index badd73b7ca1..51dd3804866 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -376,10 +376,12 @@ static int nfs4_stat_to_errno(int); decode_locku_maxsz) #define NFS4_enc_access_sz (compound_encode_hdr_maxsz + \ encode_putfh_maxsz + \ - encode_access_maxsz) + encode_access_maxsz + \ + encode_getattr_maxsz) #define NFS4_dec_access_sz (compound_decode_hdr_maxsz + \ decode_putfh_maxsz + \ - decode_access_maxsz) + decode_access_maxsz + \ + decode_getattr_maxsz) #define NFS4_enc_getattr_sz (compound_encode_hdr_maxsz + \ encode_putfh_maxsz + \ encode_getattr_maxsz) @@ -562,7 +564,6 @@ struct compound_hdr { #define RESERVE_SPACE(nbytes) do { \ p = xdr_reserve_space(xdr, nbytes); \ - if (!p) printk("RESERVE_SPACE(%d) failed in function %s\n", (int) (nbytes), __FUNCTION__); \ BUG_ON(!p); \ } while (0) @@ -628,8 +629,8 @@ static int encode_attrs(struct xdr_stream *xdr, const struct iattr *iap, const s if (iap->ia_valid & ATTR_UID) { owner_namelen = nfs_map_uid_to_name(server->nfs_client, iap->ia_uid, owner_name); if (owner_namelen < 0) { - printk(KERN_WARNING "nfs: couldn't resolve uid %d to string\n", - iap->ia_uid); + dprintk("nfs: couldn't resolve uid %d to string\n", + iap->ia_uid); /* XXX */ strcpy(owner_name, "nobody"); owner_namelen = sizeof("nobody") - 1; @@ -640,8 +641,8 @@ static int encode_attrs(struct xdr_stream *xdr, const struct iattr *iap, const s if (iap->ia_valid & ATTR_GID) { owner_grouplen = nfs_map_gid_to_group(server->nfs_client, iap->ia_gid, owner_group); if (owner_grouplen < 0) { - printk(KERN_WARNING "nfs4: couldn't resolve gid %d to string\n", - iap->ia_gid); + dprintk("nfs: couldn't resolve gid %d to string\n", + iap->ia_gid); strcpy(owner_group, "nobody"); owner_grouplen = sizeof("nobody") - 1; /* goto out; */ @@ -711,7 +712,7 @@ static int encode_attrs(struct xdr_stream *xdr, const struct iattr *iap, const s * Now we backfill the bitmap and the attribute buffer length. */ if (len != ((char *)p - (char *)q) + 4) { - printk ("encode_attr: Attr length calculation error! %u != %Zu\n", + printk(KERN_ERR "nfs: Attr length error, %u != %Zu\n", len, ((char *)p - (char *)q) + 4); BUG(); } @@ -1376,14 +1377,20 @@ static int nfs4_xdr_enc_access(struct rpc_rqst *req, __be32 *p, const struct nfs { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 2, + .nops = 3, }; int status; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); - if ((status = encode_putfh(&xdr, args->fh)) == 0) - status = encode_access(&xdr, args->access); + status = encode_putfh(&xdr, args->fh); + if (status != 0) + goto out; + status = encode_access(&xdr, args->access); + if (status != 0) + goto out; + status = encode_getfattr(&xdr, args->bitmask); +out: return status; } @@ -1857,6 +1864,7 @@ static int nfs4_xdr_enc_read(struct rpc_rqst *req, __be32 *p, struct nfs_readarg replen = (RPC_REPHDRSIZE + auth->au_rslack + NFS4_dec_read_sz) << 2; xdr_inline_pages(&req->rq_rcv_buf, replen, args->pages, args->pgbase, args->count); + req->rq_rcv_buf.flags |= XDRBUF_READ; out: return status; } @@ -1933,6 +1941,7 @@ static int nfs4_xdr_enc_write(struct rpc_rqst *req, __be32 *p, struct nfs_writea status = encode_write(&xdr, args); if (status) goto out; + req->rq_snd_buf.flags |= XDRBUF_WRITE; status = encode_getfattr(&xdr, args->bitmask); out: return status; @@ -2180,9 +2189,9 @@ out: #define READ_BUF(nbytes) do { \ p = xdr_inline_decode(xdr, nbytes); \ if (unlikely(!p)) { \ - printk(KERN_INFO "%s: prematurely hit end of receive" \ + dprintk("nfs: %s: prematurely hit end of receive" \ " buffer\n", __FUNCTION__); \ - printk(KERN_INFO "%s: xdr->p=%p, bytes=%u, xdr->end=%p\n", \ + dprintk("nfs: %s: xdr->p=%p, bytes=%u, xdr->end=%p\n", \ __FUNCTION__, xdr->p, nbytes, xdr->end); \ return -EIO; \ } \ @@ -2223,9 +2232,8 @@ static int decode_op_hdr(struct xdr_stream *xdr, enum nfs_opnum4 expected) READ_BUF(8); READ32(opnum); if (opnum != expected) { - printk(KERN_NOTICE - "nfs4_decode_op_hdr: Server returned operation" - " %d but we issued a request for %d\n", + dprintk("nfs: Server returned operation" + " %d but we issued a request for %d\n", opnum, expected); return -EIO; } @@ -2758,7 +2766,7 @@ static int decode_attr_owner(struct xdr_stream *xdr, uint32_t *bitmap, struct nf dprintk("%s: nfs_map_name_to_uid failed!\n", __FUNCTION__); } else - printk(KERN_WARNING "%s: name too long (%u)!\n", + dprintk("%s: name too long (%u)!\n", __FUNCTION__, len); bitmap[1] &= ~FATTR4_WORD1_OWNER; } @@ -2783,7 +2791,7 @@ static int decode_attr_group(struct xdr_stream *xdr, uint32_t *bitmap, struct nf dprintk("%s: nfs_map_group_to_gid failed!\n", __FUNCTION__); } else - printk(KERN_WARNING "%s: name too long (%u)!\n", + dprintk("%s: name too long (%u)!\n", __FUNCTION__, len); bitmap[1] &= ~FATTR4_WORD1_OWNER_GROUP; } @@ -2950,7 +2958,8 @@ static int verify_attr_len(struct xdr_stream *xdr, __be32 *savep, uint32_t attrl unsigned int nwords = xdr->p - savep; if (unlikely(attrwords != nwords)) { - printk(KERN_WARNING "%s: server returned incorrect attribute length: %u %c %u\n", + dprintk("%s: server returned incorrect attribute length: " + "%u %c %u\n", __FUNCTION__, attrwords << 2, (attrwords < nwords) ? '<' : '>', @@ -3451,7 +3460,7 @@ static int decode_read(struct xdr_stream *xdr, struct rpc_rqst *req, struct nfs_ hdrlen = (u8 *) p - (u8 *) iov->iov_base; recvd = req->rq_rcv_buf.len - hdrlen; if (count > recvd) { - printk(KERN_WARNING "NFS: server cheating in read reply: " + dprintk("NFS: server cheating in read reply: " "count %u > recvd %u\n", count, recvd); count = recvd; eof = 0; @@ -3500,7 +3509,8 @@ static int decode_readdir(struct xdr_stream *xdr, struct rpc_rqst *req, struct n p += 2; /* cookie */ len = ntohl(*p++); /* filename length */ if (len > NFS4_MAXNAMLEN) { - printk(KERN_WARNING "NFS: giant filename in readdir (len 0x%x)\n", len); + dprintk("NFS: giant filename in readdir (len 0x%x)\n", + len); goto err_unmap; } xlen = XDR_QUADLEN(len); @@ -3528,7 +3538,7 @@ short_pkt: entry[0] = entry[1] = 0; /* truncate listing ? */ if (!nr) { - printk(KERN_NOTICE "NFS: readdir reply truncated!\n"); + dprintk("NFS: readdir reply truncated!\n"); entry[1] = 1; } goto out; @@ -3554,13 +3564,13 @@ static int decode_readlink(struct xdr_stream *xdr, struct rpc_rqst *req) READ_BUF(4); READ32(len); if (len >= rcvbuf->page_len || len <= 0) { - dprintk(KERN_WARNING "nfs: server returned giant symlink!\n"); + dprintk("nfs: server returned giant symlink!\n"); return -ENAMETOOLONG; } hdrlen = (char *) xdr->p - (char *) iov->iov_base; recvd = req->rq_rcv_buf.len - hdrlen; if (recvd < len) { - printk(KERN_WARNING "NFS: server cheating in readlink reply: " + dprintk("NFS: server cheating in readlink reply: " "count %u > recvd %u\n", len, recvd); return -EIO; } @@ -3643,7 +3653,7 @@ static int decode_getacl(struct xdr_stream *xdr, struct rpc_rqst *req, hdrlen = (u8 *)xdr->p - (u8 *)iov->iov_base; recvd = req->rq_rcv_buf.len - hdrlen; if (attrlen > recvd) { - printk(KERN_WARNING "NFS: server cheating in getattr" + dprintk("NFS: server cheating in getattr" " acl reply: attrlen %u > recvd %u\n", attrlen, recvd); return -EINVAL; @@ -3688,8 +3698,7 @@ static int decode_setclientid(struct xdr_stream *xdr, struct nfs_client *clp) READ_BUF(8); READ32(opnum); if (opnum != OP_SETCLIENTID) { - printk(KERN_NOTICE - "nfs4_decode_setclientid: Server returned operation" + dprintk("nfs: decode_setclientid: Server returned operation" " %d\n", opnum); return -EIO; } @@ -3783,8 +3792,13 @@ static int nfs4_xdr_dec_access(struct rpc_rqst *rqstp, __be32 *p, struct nfs4_ac xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); if ((status = decode_compound_hdr(&xdr, &hdr)) != 0) goto out; - if ((status = decode_putfh(&xdr)) == 0) - status = decode_access(&xdr, res); + status = decode_putfh(&xdr); + if (status != 0) + goto out; + status = decode_access(&xdr, res); + if (status != 0) + goto out; + decode_getfattr(&xdr, res->fattr, res->server); out: return status; } diff --git a/fs/nfs/nfsroot.c b/fs/nfs/nfsroot.c index 3490322d114..e87b44ee9ac 100644 --- a/fs/nfs/nfsroot.c +++ b/fs/nfs/nfsroot.c @@ -76,6 +76,7 @@ #include <linux/fs.h> #include <linux/init.h> #include <linux/sunrpc/clnt.h> +#include <linux/sunrpc/xprtsock.h> #include <linux/nfs.h> #include <linux/nfs_fs.h> #include <linux/nfs_mount.h> @@ -491,7 +492,7 @@ static int __init root_nfs_get_handle(void) struct sockaddr_in sin; int status; int protocol = (nfs_data.flags & NFS_MOUNT_TCP) ? - IPPROTO_TCP : IPPROTO_UDP; + XPRT_TRANSPORT_TCP : XPRT_TRANSPORT_UDP; int version = (nfs_data.flags & NFS_MOUNT_VER3) ? NFS_MNT3_VERSION : NFS_MNT_VERSION; diff --git a/fs/nfs/proc.c b/fs/nfs/proc.c index 845cdde1d8b..97669ed0550 100644 --- a/fs/nfs/proc.c +++ b/fs/nfs/proc.c @@ -476,6 +476,8 @@ nfs_proc_readdir(struct dentry *dentry, struct rpc_cred *cred, dprintk("NFS call readdir %d\n", (unsigned int)cookie); status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); + nfs_invalidate_atime(dir); + dprintk("NFS reply readdir: %d\n", status); return status; } @@ -550,6 +552,7 @@ nfs_proc_pathconf(struct nfs_server *server, struct nfs_fh *fhandle, static int nfs_read_done(struct rpc_task *task, struct nfs_read_data *data) { + nfs_invalidate_atime(data->inode); if (task->tk_status >= 0) { nfs_refresh_inode(data->inode, data->res.fattr); /* Emulate the eof flag, which isn't normally needed in NFSv2 @@ -576,7 +579,7 @@ static void nfs_proc_read_setup(struct nfs_read_data *data) static int nfs_write_done(struct rpc_task *task, struct nfs_write_data *data) { if (task->tk_status >= 0) - nfs_post_op_update_inode(data->inode, data->res.fattr); + nfs_post_op_update_inode_force_wcc(data->inode, data->res.fattr); return 0; } diff --git a/fs/nfs/read.c b/fs/nfs/read.c index 19e05633f4e..4587a86adaa 100644 --- a/fs/nfs/read.c +++ b/fs/nfs/read.c @@ -341,9 +341,6 @@ int nfs_readpage_result(struct rpc_task *task, struct nfs_read_data *data) set_bit(NFS_INO_STALE, &NFS_FLAGS(data->inode)); nfs_mark_for_revalidate(data->inode); } - spin_lock(&data->inode->i_lock); - NFS_I(data->inode)->cache_validity |= NFS_INO_INVALID_ATIME; - spin_unlock(&data->inode->i_lock); return 0; } @@ -497,8 +494,7 @@ int nfs_readpage(struct file *file, struct page *page) if (ctx == NULL) goto out_unlock; } else - ctx = get_nfs_open_context((struct nfs_open_context *) - file->private_data); + ctx = get_nfs_open_context(nfs_file_open_context(file)); error = nfs_readpage_async(ctx, inode, page); @@ -576,8 +572,7 @@ int nfs_readpages(struct file *filp, struct address_space *mapping, if (desc.ctx == NULL) return -EBADF; } else - desc.ctx = get_nfs_open_context((struct nfs_open_context *) - filp->private_data); + desc.ctx = get_nfs_open_context(nfs_file_open_context(filp)); if (rsize < PAGE_CACHE_SIZE) nfs_pageio_init(&pgio, inode, nfs_pagein_multi, rsize, 0); else diff --git a/fs/nfs/super.c b/fs/nfs/super.c index b878528b64c..fa517ae9207 100644 --- a/fs/nfs/super.c +++ b/fs/nfs/super.c @@ -33,6 +33,8 @@ #include <linux/sunrpc/clnt.h> #include <linux/sunrpc/stats.h> #include <linux/sunrpc/metrics.h> +#include <linux/sunrpc/xprtsock.h> +#include <linux/sunrpc/xprtrdma.h> #include <linux/nfs_fs.h> #include <linux/nfs_mount.h> #include <linux/nfs4_mount.h> @@ -58,36 +60,6 @@ #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, @@ -97,7 +69,7 @@ enum { Opt_ac, Opt_noac, Opt_lock, Opt_nolock, Opt_v2, Opt_v3, - Opt_udp, Opt_tcp, + Opt_udp, Opt_tcp, Opt_rdma, Opt_acl, Opt_noacl, Opt_rdirplus, Opt_nordirplus, Opt_sharecache, Opt_nosharecache, @@ -116,7 +88,7 @@ enum { /* Mount options that take string arguments */ Opt_sec, Opt_proto, Opt_mountproto, - Opt_addr, Opt_mounthost, Opt_clientaddr, + Opt_addr, Opt_mountaddr, Opt_clientaddr, /* Mount options that are ignored */ Opt_userspace, Opt_deprecated, @@ -143,6 +115,7 @@ static match_table_t nfs_mount_option_tokens = { { Opt_v3, "v3" }, { Opt_udp, "udp" }, { Opt_tcp, "tcp" }, + { Opt_rdma, "rdma" }, { Opt_acl, "acl" }, { Opt_noacl, "noacl" }, { Opt_rdirplus, "rdirplus" }, @@ -175,13 +148,14 @@ static match_table_t nfs_mount_option_tokens = { { Opt_mountproto, "mountproto=%s" }, { Opt_addr, "addr=%s" }, { Opt_clientaddr, "clientaddr=%s" }, - { Opt_mounthost, "mounthost=%s" }, + { Opt_userspace, "mounthost=%s" }, + { Opt_mountaddr, "mountaddr=%s" }, { Opt_err, NULL } }; enum { - Opt_xprt_udp, Opt_xprt_tcp, + Opt_xprt_udp, Opt_xprt_tcp, Opt_xprt_rdma, Opt_xprt_err }; @@ -189,6 +163,7 @@ enum { static match_table_t nfs_xprt_protocol_tokens = { { Opt_xprt_udp, "udp" }, { Opt_xprt_tcp, "tcp" }, + { Opt_xprt_rdma, "rdma" }, { Opt_xprt_err, NULL } }; @@ -449,7 +424,7 @@ static void nfs_show_mount_options(struct seq_file *m, struct nfs_server *nfss, const char *nostr; } nfs_info[] = { { NFS_MOUNT_SOFT, ",soft", ",hard" }, - { NFS_MOUNT_INTR, ",intr", "" }, + { NFS_MOUNT_INTR, ",intr", ",nointr" }, { NFS_MOUNT_NOCTO, ",nocto", "" }, { NFS_MOUNT_NOAC, ",noac", "" }, { NFS_MOUNT_NONLM, ",nolock", "" }, @@ -460,8 +435,6 @@ static void nfs_show_mount_options(struct seq_file *m, struct nfs_server *nfss, }; const struct proc_nfs_info *nfs_infop; struct nfs_client *clp = nfss->nfs_client; - char buf[12]; - const char *proto; seq_printf(m, ",vers=%d", clp->rpc_ops->version); seq_printf(m, ",rsize=%d", nfss->rsize); @@ -480,18 +453,8 @@ static void nfs_show_mount_options(struct seq_file *m, struct nfs_server *nfss, else seq_puts(m, nfs_infop->nostr); } - switch (nfss->client->cl_xprt->prot) { - case IPPROTO_TCP: - proto = "tcp"; - break; - case IPPROTO_UDP: - proto = "udp"; - break; - default: - snprintf(buf, sizeof(buf), "%u", nfss->client->cl_xprt->prot); - proto = buf; - } - seq_printf(m, ",proto=%s", proto); + seq_printf(m, ",proto=%s", + rpc_peeraddr2str(nfss->client, RPC_DISPLAY_PROTO)); seq_printf(m, ",timeo=%lu", 10U * clp->retrans_timeo / HZ); seq_printf(m, ",retrans=%u", clp->retrans_count); seq_printf(m, ",sec=%s", nfs_pseudoflavour_to_name(nfss->client->cl_auth->au_flavor)); @@ -506,8 +469,8 @@ static int nfs_show_options(struct seq_file *m, struct vfsmount *mnt) nfs_show_mount_options(m, nfss, 0); - seq_puts(m, ",addr="); - seq_escape(m, nfss->nfs_client->cl_hostname, " \t\n\\"); + seq_printf(m, ",addr="NIPQUAD_FMT, + NIPQUAD(nfss->nfs_client->cl_addr.sin_addr)); return 0; } @@ -698,13 +661,19 @@ static int nfs_parse_mount_options(char *raw, break; case Opt_udp: mnt->flags &= ~NFS_MOUNT_TCP; - mnt->nfs_server.protocol = IPPROTO_UDP; + mnt->nfs_server.protocol = XPRT_TRANSPORT_UDP; mnt->timeo = 7; mnt->retrans = 5; break; case Opt_tcp: mnt->flags |= NFS_MOUNT_TCP; - mnt->nfs_server.protocol = IPPROTO_TCP; + mnt->nfs_server.protocol = XPRT_TRANSPORT_TCP; + mnt->timeo = 600; + mnt->retrans = 2; + break; + case Opt_rdma: + mnt->flags |= NFS_MOUNT_TCP; /* for side protocols */ + mnt->nfs_server.protocol = XPRT_TRANSPORT_RDMA; mnt->timeo = 600; mnt->retrans = 2; break; @@ -913,13 +882,20 @@ static int nfs_parse_mount_options(char *raw, switch (token) { case Opt_xprt_udp: mnt->flags &= ~NFS_MOUNT_TCP; - mnt->nfs_server.protocol = IPPROTO_UDP; + mnt->nfs_server.protocol = XPRT_TRANSPORT_UDP; mnt->timeo = 7; mnt->retrans = 5; break; case Opt_xprt_tcp: mnt->flags |= NFS_MOUNT_TCP; - mnt->nfs_server.protocol = IPPROTO_TCP; + mnt->nfs_server.protocol = XPRT_TRANSPORT_TCP; + mnt->timeo = 600; + mnt->retrans = 2; + break; + case Opt_xprt_rdma: + /* vector side protocols to TCP */ + mnt->flags |= NFS_MOUNT_TCP; + mnt->nfs_server.protocol = XPRT_TRANSPORT_RDMA; mnt->timeo = 600; mnt->retrans = 2; break; @@ -937,11 +913,12 @@ static int nfs_parse_mount_options(char *raw, switch (token) { case Opt_xprt_udp: - mnt->mount_server.protocol = IPPROTO_UDP; + mnt->mount_server.protocol = XPRT_TRANSPORT_UDP; break; case Opt_xprt_tcp: - mnt->mount_server.protocol = IPPROTO_TCP; + mnt->mount_server.protocol = XPRT_TRANSPORT_TCP; break; + case Opt_xprt_rdma: /* not used for side protocols */ default: goto out_unrec_xprt; } @@ -961,7 +938,7 @@ static int nfs_parse_mount_options(char *raw, goto out_nomem; mnt->client_address = string; break; - case Opt_mounthost: + case Opt_mountaddr: string = match_strdup(args); if (string == NULL) goto out_nomem; @@ -1027,16 +1004,10 @@ static int nfs_try_mount(struct nfs_parsed_mount_data *args, sin = args->mount_server.address; else 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); + /* + * autobind will be used if mount_server.port == 0 + */ + sin.sin_port = htons(args->mount_server.port); /* * Now ask the mount server to map our export path @@ -1049,14 +1020,11 @@ static int nfs_try_mount(struct nfs_parsed_mount_data *args, args->mount_server.version, args->mount_server.protocol, root_fh); - if (status < 0) - goto out_err; - - return status; + if (status == 0) + return 0; -out_err: - dfprintk(MOUNT, "NFS: unable to contact server on host " - NIPQUAD_FMT "\n", NIPQUAD(sin.sin_addr.s_addr)); + dfprintk(MOUNT, "NFS: unable to mount server " NIPQUAD_FMT + ", error %d\n", NIPQUAD(sin.sin_addr.s_addr), status); return status; } @@ -1079,15 +1047,31 @@ out_err: * 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, +static int nfs_validate_mount_data(void *options, + struct nfs_parsed_mount_data *args, struct nfs_fh *mntfh, const char *dev_name) { - struct nfs_mount_data *data = *options; + struct nfs_mount_data *data = (struct nfs_mount_data *)options; if (data == NULL) goto out_no_data; + memset(args, 0, sizeof(*args)); + args->flags = (NFS_MOUNT_VER3 | NFS_MOUNT_TCP); + args->rsize = NFS_MAX_FILE_IO_SIZE; + args->wsize = NFS_MAX_FILE_IO_SIZE; + args->timeo = 600; + args->retrans = 2; + args->acregmin = 3; + args->acregmax = 60; + args->acdirmin = 30; + args->acdirmax = 60; + args->mount_server.protocol = XPRT_TRANSPORT_UDP; + args->mount_server.program = NFS_MNT_PROGRAM; + args->nfs_server.protocol = XPRT_TRANSPORT_TCP; + args->nfs_server.program = NFS_PROGRAM; + switch (data->version) { case 1: data->namlen = 0; @@ -1116,92 +1100,73 @@ static int nfs_validate_mount_data(struct nfs_mount_data **options, if (mntfh->size < sizeof(mntfh->data)) memset(mntfh->data + mntfh->size, 0, sizeof(mntfh->data) - mntfh->size); + + if (!nfs_verify_server_address((struct sockaddr *) &data->addr)) + goto out_no_address; + + /* + * Translate to nfs_parsed_mount_data, which nfs_fill_super + * can deal with. + */ + args->flags = data->flags; + args->rsize = data->rsize; + args->wsize = data->wsize; + args->flags = data->flags; + args->timeo = data->timeo; + args->retrans = data->retrans; + args->acregmin = data->acregmin; + args->acregmax = data->acregmax; + args->acdirmin = data->acdirmin; + args->acdirmax = data->acdirmax; + args->nfs_server.address = data->addr; + if (!(data->flags & NFS_MOUNT_TCP)) + args->nfs_server.protocol = XPRT_TRANSPORT_UDP; + /* N.B. caller will free nfs_server.hostname in all cases */ + args->nfs_server.hostname = kstrdup(data->hostname, GFP_KERNEL); + args->namlen = data->namlen; + args->bsize = data->bsize; + args->auth_flavors[0] = data->pseudoflavor; 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; + if (nfs_parse_mount_options((char *)options, args) == 0) + return -EINVAL; - /* - * NB: after this point, caller will free "data" - * if we return an error - */ - *options = data; + if (!nfs_verify_server_address((struct sockaddr *) + &args->nfs_server.address)) + goto out_no_address; c = strchr(dev_name, ':'); if (c == NULL) return -EINVAL; len = c - dev_name; - if (len > sizeof(data->hostname)) - return -ENAMETOOLONG; - strncpy(data->hostname, dev_name, len); - args.nfs_server.hostname = data->hostname; + /* N.B. caller will free nfs_server.hostname in all cases */ + args->nfs_server.hostname = kstrndup(dev_name, len, GFP_KERNEL); c++; if (strlen(c) > NFS_MAXPATHLEN) return -ENAMETOOLONG; - args.nfs_server.export_path = c; + args->nfs_server.export_path = c; - status = nfs_try_mount(&args, mntfh); + status = nfs_try_mount(args, mntfh); if (status) return status; - /* - * 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; } } - if (!(data->flags & NFS_MOUNT_SECFLAVOUR)) - data->pseudoflavor = RPC_AUTH_UNIX; + if (!(args->flags & NFS_MOUNT_SECFLAVOUR)) + args->auth_flavors[0] = RPC_AUTH_UNIX; #ifndef CONFIG_NFS_V3 - if (data->flags & NFS_MOUNT_VER3) + if (args->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: @@ -1258,7 +1223,8 @@ static inline void nfs_initialise_sb(struct super_block *sb) /* * Finish setting up an NFS2/3 superblock */ -static void nfs_fill_super(struct super_block *sb, struct nfs_mount_data *data) +static void nfs_fill_super(struct super_block *sb, + struct nfs_parsed_mount_data *data) { struct nfs_server *server = NFS_SB(sb); @@ -1379,7 +1345,7 @@ static int nfs_get_sb(struct file_system_type *fs_type, struct nfs_server *server = NULL; struct super_block *s; struct nfs_fh mntfh; - struct nfs_mount_data *data = raw_data; + struct nfs_parsed_mount_data data; struct dentry *mntroot; int (*compare_super)(struct super_block *, void *) = nfs_compare_super; struct nfs_sb_mountdata sb_mntdata = { @@ -1388,12 +1354,12 @@ static int nfs_get_sb(struct file_system_type *fs_type, int error; /* Validate the mount data */ - error = nfs_validate_mount_data(&data, &mntfh, dev_name); + error = nfs_validate_mount_data(raw_data, &data, &mntfh, dev_name); if (error < 0) goto out; /* Get a volume representation */ - server = nfs_create_server(data, &mntfh); + server = nfs_create_server(&data, &mntfh); if (IS_ERR(server)) { error = PTR_ERR(server); goto out; @@ -1417,7 +1383,7 @@ static int nfs_get_sb(struct file_system_type *fs_type, if (!s->s_root) { /* initial superblock/root creation */ - nfs_fill_super(s, data); + nfs_fill_super(s, &data); } mntroot = nfs_get_root(s, &mntfh); @@ -1432,8 +1398,7 @@ static int nfs_get_sb(struct file_system_type *fs_type, error = 0; out: - if (data != raw_data) - kfree(data); + kfree(data.nfs_server.hostname); return error; out_err_nosb: @@ -1559,38 +1524,49 @@ static void nfs4_fill_super(struct super_block *sb) /* * 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) +static int nfs4_validate_mount_data(void *options, + struct nfs_parsed_mount_data *args, + const char *dev_name) { - struct nfs4_mount_data *data = *options; + struct nfs4_mount_data *data = (struct nfs4_mount_data *)options; char *c; if (data == NULL) goto out_no_data; + memset(args, 0, sizeof(*args)); + args->rsize = NFS_MAX_FILE_IO_SIZE; + args->wsize = NFS_MAX_FILE_IO_SIZE; + args->timeo = 600; + args->retrans = 2; + args->acregmin = 3; + args->acregmax = 60; + args->acdirmin = 30; + args->acdirmax = 60; + args->nfs_server.protocol = XPRT_TRANSPORT_TCP; + switch (data->version) { case 1: - if (data->host_addrlen != sizeof(*addr)) + if (data->host_addrlen != sizeof(args->nfs_server.address)) goto out_no_address; - if (copy_from_user(addr, data->host_addr, sizeof(*addr))) + if (copy_from_user(&args->nfs_server.address, + data->host_addr, + sizeof(args->nfs_server.address))) return -EFAULT; - if (addr->sin_port == 0) - addr->sin_port = htons(NFS_PORT); - if (!nfs_verify_server_address((struct sockaddr *) addr)) + if (args->nfs_server.address.sin_port == 0) + args->nfs_server.address.sin_port = htons(NFS_PORT); + if (!nfs_verify_server_address((struct sockaddr *) + &args->nfs_server.address)) goto out_no_address; switch (data->auth_flavourlen) { case 0: - *authflavour = RPC_AUTH_UNIX; + args->auth_flavors[0] = RPC_AUTH_UNIX; break; case 1: - if (copy_from_user(authflavour, data->auth_flavours, - sizeof(*authflavour))) + if (copy_from_user(&args->auth_flavors[0], + data->auth_flavours, + sizeof(args->auth_flavors[0]))) return -EFAULT; break; default: @@ -1600,75 +1576,57 @@ static int nfs4_validate_mount_data(struct nfs4_mount_data **options, c = strndup_user(data->hostname.data, NFS4_MAXNAMLEN); if (IS_ERR(c)) return PTR_ERR(c); - *hostname = c; + args->nfs_server.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); + args->nfs_server.export_path = c; + dfprintk(MOUNT, "NFS: MNTPATH: '%s'\n", c); c = strndup_user(data->client_addr.data, 16); if (IS_ERR(c)) return PTR_ERR(c); - *ip_addr = c; + args->client_address = c; + + /* + * Translate to nfs_parsed_mount_data, which nfs4_fill_super + * can deal with. + */ + + args->flags = data->flags & NFS4_MOUNT_FLAGMASK; + args->rsize = data->rsize; + args->wsize = data->wsize; + args->timeo = data->timeo; + args->retrans = data->retrans; + args->acregmin = data->acregmin; + args->acregmax = data->acregmax; + args->acdirmin = data->acdirmin; + args->acdirmax = data->acdirmax; + args->nfs_server.protocol = data->proto; 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) + + if (nfs_parse_mount_options((char *)options, args) == 0) return -EINVAL; if (!nfs_verify_server_address((struct sockaddr *) - &args.nfs_server.address)) + &args->nfs_server.address)) return -EINVAL; - *addr = args.nfs_server.address; - switch (args.auth_flavor_len) { + switch (args->auth_flavor_len) { case 0: - *authflavour = RPC_AUTH_UNIX; + args->auth_flavors[0] = 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, ':'); @@ -1678,27 +1636,25 @@ static int nfs4_validate_mount_data(struct nfs4_mount_data **options, len = c - dev_name; if (len > NFS4_MAXNAMLEN) return -ENAMETOOLONG; - *hostname = kzalloc(len, GFP_KERNEL); - if (*hostname == NULL) + args->nfs_server.hostname = kzalloc(len, GFP_KERNEL); + if (args->nfs_server.hostname == NULL) return -ENOMEM; - strncpy(*hostname, dev_name, len - 1); + strncpy(args->nfs_server.hostname, dev_name, len - 1); c++; /* step over the ':' */ len = strlen(c); if (len > NFS4_MAXPATHLEN) return -ENAMETOOLONG; - *mntpath = kzalloc(len + 1, GFP_KERNEL); - if (*mntpath == NULL) + args->nfs_server.export_path = kzalloc(len + 1, GFP_KERNEL); + if (args->nfs_server.export_path == NULL) return -ENOMEM; - strncpy(*mntpath, c, len); + strncpy(args->nfs_server.export_path, c, len); - dprintk("MNTPATH: %s\n", *mntpath); + dprintk("MNTPATH: %s\n", args->nfs_server.export_path); - if (args.client_address == NULL) + if (args->client_address == NULL) goto out_no_client_address; - *ip_addr = args.client_address; - break; } } @@ -1729,14 +1685,11 @@ out_no_client_address: static int nfs4_get_sb(struct file_system_type *fs_type, int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt) { - struct nfs4_mount_data *data = raw_data; + struct nfs_parsed_mount_data data; struct super_block *s; struct nfs_server *server; - struct sockaddr_in addr; - rpc_authflavor_t authflavour; struct nfs_fh mntfh; struct dentry *mntroot; - char *mntpath = NULL, *hostname = NULL, *ip_addr = NULL; int (*compare_super)(struct super_block *, void *) = nfs_compare_super; struct nfs_sb_mountdata sb_mntdata = { .mntflags = flags, @@ -1744,14 +1697,12 @@ static int nfs4_get_sb(struct file_system_type *fs_type, int error; /* Validate the mount data */ - error = nfs4_validate_mount_data(&data, dev_name, &addr, &authflavour, - &hostname, &mntpath, &ip_addr); + error = nfs4_validate_mount_data(raw_data, &data, dev_name); if (error < 0) goto out; /* Get a volume representation */ - server = nfs4_create_server(data, hostname, &addr, mntpath, ip_addr, - authflavour, &mntfh); + server = nfs4_create_server(&data, &mntfh); if (IS_ERR(server)) { error = PTR_ERR(server); goto out; @@ -1790,9 +1741,9 @@ static int nfs4_get_sb(struct file_system_type *fs_type, error = 0; out: - kfree(ip_addr); - kfree(mntpath); - kfree(hostname); + kfree(data.client_address); + kfree(data.nfs_server.export_path); + kfree(data.nfs_server.hostname); return error; out_free: diff --git a/fs/nfs/unlink.c b/fs/nfs/unlink.c index 045ab805c17..1aed850d18f 100644 --- a/fs/nfs/unlink.c +++ b/fs/nfs/unlink.c @@ -66,7 +66,6 @@ static void nfs_async_unlink_init(struct rpc_task *task, void *calldata) .rpc_cred = data->cred, }; - nfs_begin_data_update(dir); NFS_PROTO(dir)->unlink_setup(&msg, dir); rpc_call_setup(task, &msg, 0); } @@ -84,8 +83,6 @@ static void nfs_async_unlink_done(struct rpc_task *task, void *calldata) if (!NFS_PROTO(dir)->unlink_done(task, dir)) rpc_restart_call(task); - else - nfs_end_data_update(dir); } /** diff --git a/fs/nfs/write.c b/fs/nfs/write.c index 0d7a77cc394..e2bb66c3440 100644 --- a/fs/nfs/write.c +++ b/fs/nfs/write.c @@ -110,6 +110,13 @@ void nfs_writedata_release(void *wdata) nfs_writedata_free(wdata); } +static void nfs_context_set_write_error(struct nfs_open_context *ctx, int error) +{ + ctx->error = error; + smp_wmb(); + set_bit(NFS_CONTEXT_ERROR_WRITE, &ctx->flags); +} + static struct nfs_page *nfs_page_find_request_locked(struct page *page) { struct nfs_page *req = NULL; @@ -243,10 +250,7 @@ static void nfs_end_page_writeback(struct page *page) /* * Find an associated nfs write request, and prepare to flush it out - * Returns 1 if there was no write request, or if the request was - * already tagged by nfs_set_page_dirty.Returns 0 if the request - * was not tagged. - * May also return an error if the user signalled nfs_wait_on_request(). + * May return an error if the user signalled nfs_wait_on_request(). */ static int nfs_page_async_flush(struct nfs_pageio_descriptor *pgio, struct page *page) @@ -261,7 +265,7 @@ static int nfs_page_async_flush(struct nfs_pageio_descriptor *pgio, req = nfs_page_find_request_locked(page); if (req == NULL) { spin_unlock(&inode->i_lock); - return 1; + return 0; } if (nfs_lock_request_dontget(req)) break; @@ -282,7 +286,7 @@ static int nfs_page_async_flush(struct nfs_pageio_descriptor *pgio, spin_unlock(&inode->i_lock); nfs_unlock_request(req); nfs_pageio_complete(pgio); - return 1; + return 0; } if (nfs_set_page_writeback(page) != 0) { spin_unlock(&inode->i_lock); @@ -290,70 +294,56 @@ static int nfs_page_async_flush(struct nfs_pageio_descriptor *pgio, } radix_tree_tag_set(&nfsi->nfs_page_tree, req->wb_index, NFS_PAGE_TAG_LOCKED); - ret = test_bit(PG_NEED_FLUSH, &req->wb_flags); spin_unlock(&inode->i_lock); nfs_pageio_add_request(pgio, req); - return ret; + return 0; } -/* - * Write an mmapped page to the server. - */ -static int nfs_writepage_locked(struct page *page, struct writeback_control *wbc) +static int nfs_do_writepage(struct page *page, struct writeback_control *wbc, struct nfs_pageio_descriptor *pgio) { - struct nfs_pageio_descriptor mypgio, *pgio; - struct nfs_open_context *ctx; struct inode *inode = page->mapping->host; - unsigned offset; - int err; nfs_inc_stats(inode, NFSIOS_VFSWRITEPAGE); nfs_add_stats(inode, NFSIOS_WRITEPAGES, 1); - if (wbc->for_writepages) - pgio = wbc->fs_private; - else { - nfs_pageio_init_write(&mypgio, inode, wb_priority(wbc)); - pgio = &mypgio; - } - nfs_pageio_cond_complete(pgio, page->index); + return nfs_page_async_flush(pgio, page); +} - err = nfs_page_async_flush(pgio, page); - if (err <= 0) - goto out; - err = 0; - offset = nfs_page_length(page); - if (!offset) - goto out; - - nfs_pageio_cond_complete(pgio, page->index); +/* + * Write an mmapped page to the server. + */ +static int nfs_writepage_locked(struct page *page, struct writeback_control *wbc) +{ + struct nfs_pageio_descriptor pgio; + int err; - ctx = nfs_find_open_context(inode, NULL, FMODE_WRITE); - if (ctx == NULL) { - err = -EBADF; - goto out; - } - err = nfs_writepage_setup(ctx, page, 0, offset); - put_nfs_open_context(ctx); - if (err != 0) - goto out; - err = nfs_page_async_flush(pgio, page); - if (err > 0) - err = 0; -out: - if (!wbc->for_writepages) - nfs_pageio_complete(pgio); - return err; + nfs_pageio_init_write(&pgio, page->mapping->host, wb_priority(wbc)); + err = nfs_do_writepage(page, wbc, &pgio); + nfs_pageio_complete(&pgio); + if (err < 0) + return err; + if (pgio.pg_error < 0) + return pgio.pg_error; + return 0; } int nfs_writepage(struct page *page, struct writeback_control *wbc) { - int err; + int ret; + + ret = nfs_writepage_locked(page, wbc); + unlock_page(page); + return ret; +} + +static int nfs_writepages_callback(struct page *page, struct writeback_control *wbc, void *data) +{ + int ret; - err = nfs_writepage_locked(page, wbc); + ret = nfs_do_writepage(page, wbc, data); unlock_page(page); - return err; + return ret; } int nfs_writepages(struct address_space *mapping, struct writeback_control *wbc) @@ -365,12 +355,11 @@ int nfs_writepages(struct address_space *mapping, struct writeback_control *wbc) nfs_inc_stats(inode, NFSIOS_VFSWRITEPAGES); nfs_pageio_init_write(&pgio, inode, wb_priority(wbc)); - wbc->fs_private = &pgio; - err = generic_writepages(mapping, wbc); + err = write_cache_pages(mapping, wbc, nfs_writepages_callback, &pgio); nfs_pageio_complete(&pgio); - if (err) + if (err < 0) return err; - if (pgio.pg_error) + if (pgio.pg_error < 0) return pgio.pg_error; return 0; } @@ -389,14 +378,11 @@ static int nfs_inode_add_request(struct inode *inode, struct nfs_page *req) return error; if (!nfsi->npages) { igrab(inode); - nfs_begin_data_update(inode); if (nfs_have_delegation(inode, FMODE_WRITE)) nfsi->change_attr++; } SetPagePrivate(req->wb_page); set_page_private(req->wb_page, (unsigned long)req); - if (PageDirty(req->wb_page)) - set_bit(PG_NEED_FLUSH, &req->wb_flags); nfsi->npages++; kref_get(&req->wb_kref); return 0; @@ -416,12 +402,9 @@ static void nfs_inode_remove_request(struct nfs_page *req) set_page_private(req->wb_page, 0); ClearPagePrivate(req->wb_page); radix_tree_delete(&nfsi->nfs_page_tree, req->wb_index); - if (test_and_clear_bit(PG_NEED_FLUSH, &req->wb_flags)) - __set_page_dirty_nobuffers(req->wb_page); nfsi->npages--; if (!nfsi->npages) { spin_unlock(&inode->i_lock); - nfs_end_data_update(inode); iput(inode); } else spin_unlock(&inode->i_lock); @@ -682,7 +665,7 @@ static struct nfs_page * nfs_update_request(struct nfs_open_context* ctx, int nfs_flush_incompatible(struct file *file, struct page *page) { - struct nfs_open_context *ctx = (struct nfs_open_context *)file->private_data; + struct nfs_open_context *ctx = nfs_file_open_context(file); struct nfs_page *req; int do_flush, status; /* @@ -716,7 +699,7 @@ int nfs_flush_incompatible(struct file *file, struct page *page) int nfs_updatepage(struct file *file, struct page *page, unsigned int offset, unsigned int count) { - struct nfs_open_context *ctx = (struct nfs_open_context *)file->private_data; + struct nfs_open_context *ctx = nfs_file_open_context(file); struct inode *inode = page->mapping->host; int status = 0; @@ -967,7 +950,7 @@ static void nfs_writeback_done_partial(struct rpc_task *task, void *calldata) if (task->tk_status < 0) { nfs_set_pageerror(page); - req->wb_context->error = task->tk_status; + nfs_context_set_write_error(req->wb_context, task->tk_status); dprintk(", error = %d\n", task->tk_status); goto out; } @@ -1030,7 +1013,7 @@ static void nfs_writeback_done_full(struct rpc_task *task, void *calldata) if (task->tk_status < 0) { nfs_set_pageerror(page); - req->wb_context->error = task->tk_status; + nfs_context_set_write_error(req->wb_context, task->tk_status); dprintk(", error = %d\n", task->tk_status); goto remove_request; } @@ -1244,7 +1227,7 @@ static void nfs_commit_done(struct rpc_task *task, void *calldata) req->wb_bytes, (long long)req_offset(req)); if (task->tk_status < 0) { - req->wb_context->error = task->tk_status; + nfs_context_set_write_error(req->wb_context, task->tk_status); nfs_inode_remove_request(req); dprintk(", error = %d\n", task->tk_status); goto next; @@ -1347,53 +1330,52 @@ long nfs_sync_mapping_wait(struct address_space *mapping, struct writeback_contr return ret; } -/* - * flush the inode to disk. - */ -int nfs_wb_all(struct inode *inode) +static int __nfs_write_mapping(struct address_space *mapping, struct writeback_control *wbc, int how) { - struct address_space *mapping = inode->i_mapping; - struct writeback_control wbc = { - .bdi = mapping->backing_dev_info, - .sync_mode = WB_SYNC_ALL, - .nr_to_write = LONG_MAX, - .for_writepages = 1, - .range_cyclic = 1, - }; int ret; - ret = nfs_writepages(mapping, &wbc); + ret = nfs_writepages(mapping, wbc); if (ret < 0) goto out; - ret = nfs_sync_mapping_wait(mapping, &wbc, 0); - if (ret >= 0) - return 0; + ret = nfs_sync_mapping_wait(mapping, wbc, how); + if (ret < 0) + goto out; + return 0; out: __mark_inode_dirty(mapping->host, I_DIRTY_PAGES); return ret; } -int nfs_sync_mapping_range(struct address_space *mapping, loff_t range_start, loff_t range_end, int how) +/* Two pass sync: first using WB_SYNC_NONE, then WB_SYNC_ALL */ +static int nfs_write_mapping(struct address_space *mapping, int how) { struct writeback_control wbc = { .bdi = mapping->backing_dev_info, - .sync_mode = WB_SYNC_ALL, + .sync_mode = WB_SYNC_NONE, .nr_to_write = LONG_MAX, - .range_start = range_start, - .range_end = range_end, .for_writepages = 1, + .range_cyclic = 1, }; int ret; - ret = nfs_writepages(mapping, &wbc); + ret = __nfs_write_mapping(mapping, &wbc, how); if (ret < 0) - goto out; - ret = nfs_sync_mapping_wait(mapping, &wbc, how); - if (ret >= 0) - return 0; -out: - __mark_inode_dirty(mapping->host, I_DIRTY_PAGES); - return ret; + return ret; + wbc.sync_mode = WB_SYNC_ALL; + return __nfs_write_mapping(mapping, &wbc, how); +} + +/* + * flush the inode to disk. + */ +int nfs_wb_all(struct inode *inode) +{ + return nfs_write_mapping(inode->i_mapping, 0); +} + +int nfs_wb_nocommit(struct inode *inode) +{ + return nfs_write_mapping(inode->i_mapping, FLUSH_NOCOMMIT); } int nfs_wb_page_cancel(struct inode *inode, struct page *page) @@ -1477,35 +1459,6 @@ int nfs_wb_page(struct inode *inode, struct page* page) return nfs_wb_page_priority(inode, page, FLUSH_STABLE); } -int nfs_set_page_dirty(struct page *page) -{ - struct address_space *mapping = page->mapping; - struct inode *inode; - struct nfs_page *req; - int ret; - - if (!mapping) - goto out_raced; - inode = mapping->host; - if (!inode) - goto out_raced; - 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(&inode->i_lock); - nfs_release_request(req); - return ret; - } - ret = __set_page_dirty_nobuffers(page); - spin_unlock(&inode->i_lock); - return ret; -out_raced: - return !TestSetPageDirty(page); -} - - int __init nfs_init_writepagecache(void) { nfs_wdata_cachep = kmem_cache_create("nfs_write_data", |