diff options
Diffstat (limited to 'fs/nfsd/vfs.c')
| -rw-r--r-- | fs/nfsd/vfs.c | 1143 |
1 files changed, 667 insertions, 476 deletions
diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c index 0aa02eb18bd3..f650afedd67f 100644 --- a/fs/nfsd/vfs.c +++ b/fs/nfsd/vfs.c @@ -26,20 +26,20 @@ #include <linux/xattr.h> #include <linux/jhash.h> #include <linux/ima.h> +#include <linux/pagemap.h> #include <linux/slab.h> #include <linux/uaccess.h> #include <linux/exportfs.h> #include <linux/writeback.h> #include <linux/security.h> -#ifdef CONFIG_NFSD_V3 #include "xdr3.h" -#endif /* CONFIG_NFSD_V3 */ #ifdef CONFIG_NFSD_V4 #include "../internal.h" #include "acl.h" #include "idmap.h" +#include "xdr4.h" #endif /* CONFIG_NFSD_V4 */ #include "nfsd.h" @@ -199,27 +199,13 @@ nfsd_lookup_dentry(struct svc_rqst *rqstp, struct svc_fh *fhp, goto out_nfserr; } } else { - /* - * In the nfsd4_open() case, this may be held across - * subsequent open and delegation acquisition which may - * need to take the child's i_mutex: - */ - fh_lock_nested(fhp, I_MUTEX_PARENT); - dentry = lookup_one_len(name, dparent, len); + dentry = lookup_one_len_unlocked(name, dparent, len); host_err = PTR_ERR(dentry); if (IS_ERR(dentry)) goto out_nfserr; if (nfsd_mountpoint(dentry, exp)) { - /* - * We don't need the i_mutex after all. It's - * still possible we could open this (regular - * files can be mountpoints too), but the - * i_mutex is just there to prevent renames of - * something that we might be about to delegate, - * and a mountpoint won't be renamed: - */ - fh_unlock(fhp); - if ((host_err = nfsd_cross_mnt(rqstp, &dentry, &exp))) { + host_err = nfsd_cross_mnt(rqstp, &dentry, &exp); + if (host_err) { dput(dentry); goto out_nfserr; } @@ -234,7 +220,15 @@ out_nfserr: return nfserrno(host_err); } -/* +/** + * nfsd_lookup - look up a single path component for nfsd + * + * @rqstp: the request context + * @fhp: the file handle of the directory + * @name: the component name, or %NULL to look up parent + * @len: length of name to examine + * @resfh: pointer to pre-initialised filehandle to hold result. + * * Look up one component of a pathname. * N.B. After this call _both_ fhp and resfh need an fh_put * @@ -244,11 +238,11 @@ out_nfserr: * returned. Otherwise the covered directory is returned. * NOTE: this mountpoint crossing is not supported properly by all * clients and is explicitly disallowed for NFSv3 - * NeilBrown <neilb@cse.unsw.edu.au> + * */ __be32 nfsd_lookup(struct svc_rqst *rqstp, struct svc_fh *fhp, const char *name, - unsigned int len, struct svc_fh *resfh) + unsigned int len, struct svc_fh *resfh) { struct svc_export *exp; struct dentry *dentry; @@ -306,6 +300,10 @@ commit_metadata(struct svc_fh *fhp) static void nfsd_sanitize_attrs(struct inode *inode, struct iattr *iap) { + /* Ignore mode updates on symlinks */ + if (S_ISLNK(inode->i_mode)) + iap->ia_valid &= ~ATTR_MODE; + /* sanitize the mode change */ if (iap->ia_valid & ATTR_MODE) { iap->ia_mode &= S_IALLUGO; @@ -333,7 +331,6 @@ nfsd_get_write_access(struct svc_rqst *rqstp, struct svc_fh *fhp, struct iattr *iap) { struct inode *inode = d_inode(fhp->fh_dentry); - int host_err; if (iap->ia_size < inode->i_size) { __be32 err; @@ -343,37 +340,80 @@ nfsd_get_write_access(struct svc_rqst *rqstp, struct svc_fh *fhp, if (err) return err; } + return nfserrno(get_write_access(inode)); +} - host_err = get_write_access(inode); - if (host_err) - goto out_nfserrno; +static int __nfsd_setattr(struct dentry *dentry, struct iattr *iap) +{ + int host_err; - host_err = locks_verify_truncate(inode, NULL, iap->ia_size); - if (host_err) - goto out_put_write_access; - return 0; + if (iap->ia_valid & ATTR_SIZE) { + /* + * RFC5661, Section 18.30.4: + * Changing the size of a file with SETATTR indirectly + * changes the time_modify and change attributes. + * + * (and similar for the older RFCs) + */ + struct iattr size_attr = { + .ia_valid = ATTR_SIZE | ATTR_CTIME | ATTR_MTIME, + .ia_size = iap->ia_size, + }; -out_put_write_access: - put_write_access(inode); -out_nfserrno: - return nfserrno(host_err); + if (iap->ia_size < 0) + return -EFBIG; + + host_err = notify_change(&init_user_ns, dentry, &size_attr, NULL); + if (host_err) + return host_err; + iap->ia_valid &= ~ATTR_SIZE; + + /* + * Avoid the additional setattr call below if the only other + * attribute that the client sends is the mtime, as we update + * it as part of the size change above. + */ + if ((iap->ia_valid & ~ATTR_MTIME) == 0) + return 0; + } + + if (!iap->ia_valid) + return 0; + + iap->ia_valid |= ATTR_CTIME; + return notify_change(&init_user_ns, dentry, iap, NULL); } -/* - * Set various file attributes. After this call fhp needs an fh_put. +/** + * nfsd_setattr - Set various file attributes. + * @rqstp: controlling RPC transaction + * @fhp: filehandle of target + * @attr: attributes to set + * @check_guard: set to 1 if guardtime is a valid timestamp + * @guardtime: do not act if ctime.tv_sec does not match this timestamp + * + * This call may adjust the contents of @attr (in particular, this + * call may change the bits in the na_iattr.ia_valid field). + * + * Returns nfs_ok on success, otherwise an NFS status code is + * returned. Caller must release @fhp by calling fh_put in either + * case. */ __be32 -nfsd_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp, struct iattr *iap, +nfsd_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp, + struct nfsd_attrs *attr, int check_guard, time64_t guardtime) { struct dentry *dentry; struct inode *inode; + struct iattr *iap = attr->na_iattr; int accmode = NFSD_MAY_SATTR; umode_t ftype = 0; __be32 err; int host_err; bool get_write_count; bool size_change = (iap->ia_valid & ATTR_SIZE); + int retries; if (iap->ia_valid & ATTR_SIZE) { accmode |= NFSD_MAY_WRITE|NFSD_MAY_OWNER_OVERRIDE; @@ -409,13 +449,6 @@ nfsd_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp, struct iattr *iap, dentry = fhp->fh_dentry; inode = d_inode(dentry); - /* Ignore any mode updates on symlinks */ - if (S_ISLNK(inode->i_mode)) - iap->ia_valid &= ~ATTR_MODE; - - if (!iap->ia_valid) - return 0; - nfsd_sanitize_attrs(inode, iap); if (check_guard && guardtime != inode->i_ctime.tv_sec) @@ -434,39 +467,27 @@ nfsd_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp, struct iattr *iap, return err; } - fh_lock(fhp); - if (size_change) { - /* - * RFC5661, Section 18.30.4: - * Changing the size of a file with SETATTR indirectly - * changes the time_modify and change attributes. - * - * (and similar for the older RFCs) - */ - struct iattr size_attr = { - .ia_valid = ATTR_SIZE | ATTR_CTIME | ATTR_MTIME, - .ia_size = iap->ia_size, - }; - - host_err = notify_change(dentry, &size_attr, NULL); - if (host_err) - goto out_unlock; - iap->ia_valid &= ~ATTR_SIZE; - - /* - * Avoid the additional setattr call below if the only other - * attribute that the client sends is the mtime, as we update - * it as part of the size change above. - */ - if ((iap->ia_valid & ~ATTR_MTIME) == 0) - goto out_unlock; + inode_lock(inode); + for (retries = 1;;) { + host_err = __nfsd_setattr(dentry, iap); + if (host_err != -EAGAIN || !retries--) + break; + if (!nfsd_wait_for_delegreturn(rqstp, inode)) + break; } - - iap->ia_valid |= ATTR_CTIME; - host_err = notify_change(dentry, iap, NULL); - -out_unlock: - fh_unlock(fhp); + if (attr->na_seclabel && attr->na_seclabel->len) + attr->na_labelerr = security_inode_setsecctx(dentry, + attr->na_seclabel->data, attr->na_seclabel->len); + if (IS_ENABLED(CONFIG_FS_POSIX_ACL) && attr->na_pacl) + attr->na_aclerr = set_posix_acl(&init_user_ns, + inode, ACL_TYPE_ACCESS, + attr->na_pacl); + if (IS_ENABLED(CONFIG_FS_POSIX_ACL) && + !attr->na_aclerr && attr->na_dpacl && S_ISDIR(inode->i_mode)) + attr->na_aclerr = set_posix_acl(&init_user_ns, + inode, ACL_TYPE_DEFAULT, + attr->na_dpacl); + inode_unlock(inode); if (size_change) put_write_access(inode); out: @@ -499,46 +520,29 @@ int nfsd4_is_junction(struct dentry *dentry) return 0; if (!(inode->i_mode & S_ISVTX)) return 0; - if (vfs_getxattr(dentry, NFSD_JUNCTION_XATTR_NAME, NULL, 0) <= 0) + if (vfs_getxattr(&init_user_ns, dentry, NFSD_JUNCTION_XATTR_NAME, + NULL, 0) <= 0) return 0; return 1; } -#ifdef CONFIG_NFSD_V4_SECURITY_LABEL -__be32 nfsd4_set_nfs4_label(struct svc_rqst *rqstp, struct svc_fh *fhp, - struct xdr_netobj *label) -{ - __be32 error; - int host_error; - struct dentry *dentry; - - error = fh_verify(rqstp, fhp, 0 /* S_IFREG */, NFSD_MAY_SATTR); - if (error) - return error; - dentry = fhp->fh_dentry; - - inode_lock(d_inode(dentry)); - host_error = security_inode_setsecctx(dentry, label->data, label->len); - inode_unlock(d_inode(dentry)); - return nfserrno(host_error); -} -#else -__be32 nfsd4_set_nfs4_label(struct svc_rqst *rqstp, struct svc_fh *fhp, - struct xdr_netobj *label) +static struct nfsd4_compound_state *nfsd4_get_cstate(struct svc_rqst *rqstp) { - return nfserr_notsupp; + return &((struct nfsd4_compoundres *)rqstp->rq_resp)->cstate; } -#endif -__be32 nfsd4_clone_file_range(struct nfsd_file *nf_src, u64 src_pos, - struct nfsd_file *nf_dst, u64 dst_pos, u64 count, bool sync) +__be32 nfsd4_clone_file_range(struct svc_rqst *rqstp, + struct nfsd_file *nf_src, u64 src_pos, + struct nfsd_file *nf_dst, u64 dst_pos, + u64 count, bool sync) { struct file *src = nf_src->nf_file; struct file *dst = nf_dst->nf_file; + errseq_t since; loff_t cloned; __be32 ret = 0; - down_write(&nf_dst->nf_rwsem); + since = READ_ONCE(dst->f_wb_err); cloned = vfs_clone_file_range(src, src_pos, dst, dst_pos, count, 0); if (cloned < 0) { ret = nfserrno(cloned); @@ -553,21 +557,32 @@ __be32 nfsd4_clone_file_range(struct nfsd_file *nf_src, u64 src_pos, int status = vfs_fsync_range(dst, dst_pos, dst_end, 0); if (!status) + status = filemap_check_wb_err(dst->f_mapping, since); + if (!status) status = commit_inode_metadata(file_inode(src)); if (status < 0) { - nfsd_reset_boot_verifier(net_generic(nf_dst->nf_net, - nfsd_net_id)); + struct nfsd_net *nn = net_generic(nf_dst->nf_net, + nfsd_net_id); + + trace_nfsd_clone_file_range_err(rqstp, + &nfsd4_get_cstate(rqstp)->save_fh, + src_pos, + &nfsd4_get_cstate(rqstp)->current_fh, + dst_pos, + count, status); + nfsd_reset_write_verifier(nn); + trace_nfsd_writeverf_reset(nn, rqstp, status); ret = nfserrno(status); } } out_err: - up_write(&nf_dst->nf_rwsem); return ret; } ssize_t nfsd_copy_file_range(struct file *src, u64 src_pos, struct file *dst, u64 dst_pos, u64 count) { + ssize_t ret; /* * Limit copy to 4MB to prevent indefinitely blocking an nfsd @@ -578,7 +593,12 @@ ssize_t nfsd_copy_file_range(struct file *src, u64 src_pos, struct file *dst, * limit like this and pipeline multiple COPY requests. */ count = min_t(u64, count, 1 << 22); - return vfs_copy_file_range(src, src_pos, dst, dst_pos, count, 0); + ret = vfs_copy_file_range(src, src_pos, dst, dst_pos, count, 0); + + if (ret == -EOPNOTSUPP || ret == -EXDEV) + ret = generic_copy_file_range(src, src_pos, dst, dst_pos, + count, 0); + return ret; } __be32 nfsd4_vfs_fallocate(struct svc_rqst *rqstp, struct svc_fh *fhp, @@ -598,7 +618,6 @@ __be32 nfsd4_vfs_fallocate(struct svc_rqst *rqstp, struct svc_fh *fhp, } #endif /* defined(CONFIG_NFSD_V4) */ -#ifdef CONFIG_NFSD_V3 /* * Check server access rights to a file system object */ @@ -612,6 +631,12 @@ static struct accessmap nfs3_regaccess[] = { { NFS3_ACCESS_MODIFY, NFSD_MAY_WRITE|NFSD_MAY_TRUNC }, { NFS3_ACCESS_EXTEND, NFSD_MAY_WRITE }, +#ifdef CONFIG_NFSD_V4 + { NFS4_ACCESS_XAREAD, NFSD_MAY_READ }, + { NFS4_ACCESS_XAWRITE, NFSD_MAY_WRITE }, + { NFS4_ACCESS_XALIST, NFSD_MAY_READ }, +#endif + { 0, 0 } }; @@ -622,6 +647,12 @@ static struct accessmap nfs3_diraccess[] = { { NFS3_ACCESS_EXTEND, NFSD_MAY_EXEC|NFSD_MAY_WRITE }, { NFS3_ACCESS_DELETE, NFSD_MAY_REMOVE }, +#ifdef CONFIG_NFSD_V4 + { NFS4_ACCESS_XAREAD, NFSD_MAY_READ }, + { NFS4_ACCESS_XAWRITE, NFSD_MAY_WRITE }, + { NFS4_ACCESS_XALIST, NFSD_MAY_READ }, +#endif + { 0, 0 } }; @@ -698,7 +729,6 @@ nfsd_access(struct svc_rqst *rqstp, struct svc_fh *fhp, u32 *access, u32 *suppor out: return error; } -#endif /* CONFIG_NFSD_V3 */ int nfsd_open_break_lease(struct inode *inode, int access) { @@ -731,19 +761,9 @@ __nfsd_open(struct svc_rqst *rqstp, struct svc_fh *fhp, umode_t type, path.dentry = fhp->fh_dentry; inode = d_inode(path.dentry); - /* Disallow write access to files with the append-only bit set - * or any access when mandatory locking enabled - */ err = nfserr_perm; if (IS_APPEND(inode) && (may_flags & NFSD_MAY_WRITE)) goto out; - /* - * We must ignore files (but only files) which might have mandatory - * locks on them because there is no way to know if the accesser has - * the lock. - */ - if (S_ISREG((inode)->i_mode) && mandatory_lock(inode)) - goto out; if (!inode->i_fop) goto out; @@ -788,6 +808,7 @@ nfsd_open(struct svc_rqst *rqstp, struct svc_fh *fhp, umode_t type, int may_flags, struct file **filp) { __be32 err; + bool retried = false; validate_process_creds(); /* @@ -803,21 +824,37 @@ nfsd_open(struct svc_rqst *rqstp, struct svc_fh *fhp, umode_t type, */ if (type == S_IFREG) may_flags |= NFSD_MAY_OWNER_OVERRIDE; +retry: err = fh_verify(rqstp, fhp, type, may_flags); - if (!err) + if (!err) { err = __nfsd_open(rqstp, fhp, type, may_flags, filp); + if (err == nfserr_stale && !retried) { + retried = true; + fh_put(fhp); + goto retry; + } + } validate_process_creds(); return err; } +/** + * nfsd_open_verified - Open a regular file for the filecache + * @rqstp: RPC request + * @fhp: NFS filehandle of the file to open + * @may_flags: internal permission flags + * @filp: OUT: open "struct file *" + * + * Returns an nfsstat value in network byte order. + */ __be32 -nfsd_open_verified(struct svc_rqst *rqstp, struct svc_fh *fhp, umode_t type, - int may_flags, struct file **filp) +nfsd_open_verified(struct svc_rqst *rqstp, struct svc_fh *fhp, int may_flags, + struct file **filp) { __be32 err; validate_process_creds(); - err = __nfsd_open(rqstp, fhp, type, may_flags, filp); + err = __nfsd_open(rqstp, fhp, S_IFREG, may_flags, filp); validate_process_creds(); return err; } @@ -832,28 +869,16 @@ nfsd_splice_actor(struct pipe_inode_info *pipe, struct pipe_buffer *buf, struct splice_desc *sd) { struct svc_rqst *rqstp = sd->u.data; - struct page **pp = rqstp->rq_next_page; - struct page *page = buf->page; - size_t size; - - size = sd->len; - - if (rqstp->rq_res.page_len == 0) { - get_page(page); - put_page(*rqstp->rq_next_page); - *(rqstp->rq_next_page++) = page; - rqstp->rq_res.page_base = buf->offset; - rqstp->rq_res.page_len = size; - } else if (page != pp[-1]) { - get_page(page); - if (*rqstp->rq_next_page) - put_page(*rqstp->rq_next_page); - *(rqstp->rq_next_page++) = page; - rqstp->rq_res.page_len += size; - } else - rqstp->rq_res.page_len += size; - - return size; + struct page *page = buf->page; // may be a compound one + unsigned offset = buf->offset; + + page += offset / PAGE_SIZE; + for (int i = sd->len; i > 0; i -= PAGE_SIZE) + svc_rqst_replace_page(rqstp, page++); + if (rqstp->rq_res.page_len == 0) // first call + rqstp->rq_res.page_base = offset % PAGE_SIZE; + rqstp->rq_res.page_len += sd->len; + return sd->len; } static int nfsd_direct_splice_actor(struct pipe_inode_info *pipe, @@ -877,7 +902,7 @@ static __be32 nfsd_finish_read(struct svc_rqst *rqstp, struct svc_fh *fhp, unsigned long *count, u32 *eof, ssize_t host_err) { if (host_err >= 0) { - nfsdstats.io_read += host_err; + nfsd_stats_io_read_add(fhp->fh_export, host_err); *eof = nfsd_eof_on_read(file, offset, host_err, *count); *count = host_err; fsnotify_access(file); @@ -965,26 +990,38 @@ nfsd_vfs_write(struct svc_rqst *rqstp, struct svc_fh *fhp, struct nfsd_file *nf, unsigned long *cnt, int stable, __be32 *verf) { + struct nfsd_net *nn = net_generic(SVC_NET(rqstp), nfsd_net_id); struct file *file = nf->nf_file; + struct super_block *sb = file_inode(file)->i_sb; struct svc_export *exp; struct iov_iter iter; + errseq_t since; __be32 nfserr; int host_err; int use_wgather; loff_t pos = offset; + unsigned long exp_op_flags = 0; unsigned int pflags = current->flags; rwf_t flags = 0; + bool restore_flags = false; trace_nfsd_write_opened(rqstp, fhp, offset, *cnt); - if (test_bit(RQ_LOCAL, &rqstp->rq_flags)) + if (sb->s_export_op) + exp_op_flags = sb->s_export_op->flags; + + if (test_bit(RQ_LOCAL, &rqstp->rq_flags) && + !(exp_op_flags & EXPORT_OP_REMOTE_FS)) { /* - * We want less throttling in balance_dirty_pages() - * and shrink_inactive_list() so that nfs to + * We want throttling in balance_dirty_pages() + * and shrink_inactive_list() to only consider + * the backingdev we are writing to, so that nfs to * localhost doesn't cause nfsd to lock up due to all * the client's dirty pages or its congested queue. */ - current->flags |= PF_LESS_THROTTLE; + current->flags |= PF_LOCAL_THROTTLE; + restore_flags = true; + } exp = fhp->fh_export; use_wgather = (rqstp->rq_vers == 2) && EX_WGATHER(exp); @@ -996,36 +1033,28 @@ nfsd_vfs_write(struct svc_rqst *rqstp, struct svc_fh *fhp, struct nfsd_file *nf, flags |= RWF_SYNC; iov_iter_kvec(&iter, WRITE, vec, vlen, *cnt); - if (flags & RWF_SYNC) { - down_write(&nf->nf_rwsem); - host_err = vfs_iter_write(file, &iter, &pos, flags); - if (host_err < 0) - nfsd_reset_boot_verifier(net_generic(SVC_NET(rqstp), - nfsd_net_id)); - up_write(&nf->nf_rwsem); - } else { - down_read(&nf->nf_rwsem); - if (verf) - nfsd_copy_boot_verifier(verf, - net_generic(SVC_NET(rqstp), - nfsd_net_id)); - host_err = vfs_iter_write(file, &iter, &pos, flags); - up_read(&nf->nf_rwsem); - } + since = READ_ONCE(file->f_wb_err); + if (verf) + nfsd_copy_write_verifier(verf, nn); + host_err = vfs_iter_write(file, &iter, &pos, flags); if (host_err < 0) { - nfsd_reset_boot_verifier(net_generic(SVC_NET(rqstp), - nfsd_net_id)); + nfsd_reset_write_verifier(nn); + trace_nfsd_writeverf_reset(nn, rqstp, host_err); goto out_nfserr; } *cnt = host_err; - nfsdstats.io_write += *cnt; + nfsd_stats_io_write_add(exp, *cnt); fsnotify_modify(file); + host_err = filemap_check_wb_err(file->f_mapping, since); + if (host_err < 0) + goto out_nfserr; if (stable && use_wgather) { host_err = wait_for_concurrent_writes(file); - if (host_err < 0) - nfsd_reset_boot_verifier(net_generic(SVC_NET(rqstp), - nfsd_net_id)); + if (host_err < 0) { + nfsd_reset_write_verifier(nn); + trace_nfsd_writeverf_reset(nn, rqstp, host_err); + } } out_nfserr: @@ -1036,8 +1065,8 @@ out_nfserr: trace_nfsd_write_err(rqstp, fhp, offset, host_err); nfserr = nfserrno(host_err); } - if (test_bit(RQ_LOCAL, &rqstp->rq_flags)) - current_restore_flags(pflags, PF_LESS_THROTTLE); + if (restore_flags) + current_restore_flags(pflags, PF_LOCAL_THROTTLE); return nfserr; } @@ -1099,73 +1128,105 @@ out: return err; } -#ifdef CONFIG_NFSD_V3 -/* - * Commit all pending writes to stable storage. +/** + * nfsd_commit - Commit pending writes to stable storage + * @rqstp: RPC request being processed + * @fhp: NFS filehandle + * @offset: raw offset from beginning of file + * @count: raw count of bytes to sync + * @verf: filled in with the server's current write verifier * - * Note: we only guarantee that data that lies within the range specified - * by the 'offset' and 'count' parameters will be synced. + * Note: we guarantee that data that lies within the range specified + * by the 'offset' and 'count' parameters will be synced. The server + * is permitted to sync data that lies outside this range at the + * same time. * * Unfortunately we cannot lock the file to make sure we return full WCC * data to the client, as locking happens lower down in the filesystem. + * + * Return values: + * An nfsstat value in network byte order. */ __be32 -nfsd_commit(struct svc_rqst *rqstp, struct svc_fh *fhp, - loff_t offset, unsigned long count, __be32 *verf) +nfsd_commit(struct svc_rqst *rqstp, struct svc_fh *fhp, u64 offset, + u32 count, __be32 *verf) { + u64 maxbytes; + loff_t start, end; + struct nfsd_net *nn; struct nfsd_file *nf; - loff_t end = LLONG_MAX; - __be32 err = nfserr_inval; - - if (offset < 0) - goto out; - if (count != 0) { - end = offset + (loff_t)count - 1; - if (end < offset) - goto out; - } + __be32 err; err = nfsd_file_acquire(rqstp, fhp, NFSD_MAY_WRITE|NFSD_MAY_NOT_BREAK_LEASE, &nf); if (err) goto out; + + /* + * Convert the client-provided (offset, count) range to a + * (start, end) range. If the client-provided range falls + * outside the maximum file size of the underlying FS, + * clamp the sync range appropriately. + */ + start = 0; + end = LLONG_MAX; + maxbytes = (u64)fhp->fh_dentry->d_sb->s_maxbytes; + if (offset < maxbytes) { + start = offset; + if (count && (offset + count - 1 < maxbytes)) + end = offset + count - 1; + } + + nn = net_generic(nf->nf_net, nfsd_net_id); if (EX_ISSYNC(fhp->fh_export)) { + errseq_t since = READ_ONCE(nf->nf_file->f_wb_err); int err2; - down_write(&nf->nf_rwsem); - err2 = vfs_fsync_range(nf->nf_file, offset, end, 0); + err2 = vfs_fsync_range(nf->nf_file, start, end, 0); switch (err2) { case 0: - nfsd_copy_boot_verifier(verf, net_generic(nf->nf_net, - nfsd_net_id)); + nfsd_copy_write_verifier(verf, nn); + err2 = filemap_check_wb_err(nf->nf_file->f_mapping, + since); + err = nfserrno(err2); break; case -EINVAL: err = nfserr_notsupp; break; default: + nfsd_reset_write_verifier(nn); + trace_nfsd_writeverf_reset(nn, rqstp, err2); err = nfserrno(err2); - nfsd_reset_boot_verifier(net_generic(nf->nf_net, - nfsd_net_id)); } - up_write(&nf->nf_rwsem); } else - nfsd_copy_boot_verifier(verf, net_generic(nf->nf_net, - nfsd_net_id)); + nfsd_copy_write_verifier(verf, nn); nfsd_file_put(nf); out: return err; } -#endif /* CONFIG_NFSD_V3 */ -static __be32 -nfsd_create_setattr(struct svc_rqst *rqstp, struct svc_fh *resfhp, - struct iattr *iap) +/** + * nfsd_create_setattr - Set a created file's attributes + * @rqstp: RPC transaction being executed + * @fhp: NFS filehandle of parent directory + * @resfhp: NFS filehandle of new object + * @attrs: requested attributes of new object + * + * Returns nfs_ok on success, or an nfsstat in network byte order. + */ +__be32 +nfsd_create_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp, + struct svc_fh *resfhp, struct nfsd_attrs *attrs) { + struct iattr *iap = attrs->na_iattr; + __be32 status; + /* - * Mode has already been set earlier in create: + * Mode has already been set by file creation. */ iap->ia_valid &= ~ATTR_MODE; + /* * Setting uid/gid works only for root. Irix appears to * send along the gid on create when it tries to implement @@ -1173,10 +1234,31 @@ nfsd_create_setattr(struct svc_rqst *rqstp, struct svc_fh *resfhp, */ if (!uid_eq(current_fsuid(), GLOBAL_ROOT_UID)) iap->ia_valid &= ~(ATTR_UID|ATTR_GID); + + /* + * Callers expect new file metadata to be committed even + * if the attributes have not changed. + */ if (iap->ia_valid) - return nfsd_setattr(rqstp, resfhp, iap, 0, (time64_t)0); - /* Callers expect file metadata to be committed here */ - return nfserrno(commit_metadata(resfhp)); + status = nfsd_setattr(rqstp, resfhp, attrs, 0, (time64_t)0); + else + status = nfserrno(commit_metadata(resfhp)); + + /* + * Transactional filesystems had a chance to commit changes + * for both parent and child simultaneously making the + * following commit_metadata a noop in many cases. + */ + if (!status) + status = nfserrno(commit_metadata(fhp)); + + /* + * Update the new filehandle to pick up the new attributes. + */ + if (!status) + status = fh_update(resfhp); + + return status; } /* HPUX client sometimes creates a file in mode 000, and sets size to 0. @@ -1197,26 +1279,19 @@ nfsd_check_ignore_resizing(struct iattr *iap) /* The parent directory should already be locked: */ __be32 nfsd_create_locked(struct svc_rqst *rqstp, struct svc_fh *fhp, - char *fname, int flen, struct iattr *iap, - int type, dev_t rdev, struct svc_fh *resfhp) + struct nfsd_attrs *attrs, + int type, dev_t rdev, struct svc_fh *resfhp) { struct dentry *dentry, *dchild; struct inode *dirp; + struct iattr *iap = attrs->na_iattr; __be32 err; - __be32 err2; int host_err; dentry = fhp->fh_dentry; dirp = d_inode(dentry); dchild = dget(resfhp->fh_dentry); - if (!fhp->fh_locked) { - WARN_ONCE(1, "nfsd_create: parent %pd2 not locked!\n", - dentry); - err = nfserr_io; - goto out; - } - err = nfsd_permission(rqstp, fhp->fh_export, dentry, NFSD_MAY_CREATE); if (err) goto out; @@ -1225,16 +1300,19 @@ nfsd_create_locked(struct svc_rqst *rqstp, struct svc_fh *fhp, iap->ia_mode = 0; iap->ia_mode = (iap->ia_mode & S_IALLUGO) | type; + if (!IS_POSIXACL(dirp)) + iap->ia_mode &= ~current_umask(); + err = 0; host_err = 0; switch (type) { case S_IFREG: - host_err = vfs_create(dirp, dchild, iap->ia_mode, true); + host_err = vfs_create(&init_user_ns, dirp, dchild, iap->ia_mode, true); if (!host_err) nfsd_check_ignore_resizing(iap); break; case S_IFDIR: - host_err = vfs_mkdir(dirp, dchild, iap->ia_mode); + host_err = vfs_mkdir(&init_user_ns, dirp, dchild, iap->ia_mode); if (!host_err && unlikely(d_unhashed(dchild))) { struct dentry *d; d = lookup_one_len(dchild->d_name.name, @@ -1262,7 +1340,8 @@ nfsd_create_locked(struct svc_rqst *rqstp, struct svc_fh *fhp, case S_IFBLK: case S_IFIFO: case S_IFSOCK: - host_err = vfs_mknod(dirp, dchild, iap->ia_mode, rdev); + host_err = vfs_mknod(&init_user_ns, dirp, dchild, + iap->ia_mode, rdev); break; default: printk(KERN_WARNING "nfsd: bad file type %o in nfsd_create\n", @@ -1272,22 +1351,8 @@ nfsd_create_locked(struct svc_rqst *rqstp, struct svc_fh *fhp, if (host_err < 0) goto out_nfserr; - err = nfsd_create_setattr(rqstp, resfhp, iap); + err = nfsd_create_setattr(rqstp, fhp, resfhp, attrs); - /* - * nfsd_create_setattr already committed the child. Transactional - * filesystems had a chance to commit changes for both parent and - * child simultaneously making the following commit_metadata a - * noop. - */ - err2 = nfserrno(commit_metadata(fhp)); - if (err2) - err = err2; - /* - * Update the file handle to get the new inode info. - */ - if (!err) - err = fh_update(resfhp); out: dput(dchild); return err; @@ -1305,8 +1370,8 @@ out_nfserr: */ __be32 nfsd_create(struct svc_rqst *rqstp, struct svc_fh *fhp, - char *fname, int flen, struct iattr *iap, - int type, dev_t rdev, struct svc_fh *resfhp) + char *fname, int flen, struct nfsd_attrs *attrs, + int type, dev_t rdev, struct svc_fh *resfhp) { struct dentry *dentry, *dchild = NULL; __be32 err; @@ -1325,11 +1390,13 @@ nfsd_create(struct svc_rqst *rqstp, struct svc_fh *fhp, if (host_err) return nfserrno(host_err); - fh_lock_nested(fhp, I_MUTEX_PARENT); + inode_lock_nested(dentry->d_inode, I_MUTEX_PARENT); dchild = lookup_one_len(fname, dentry, flen); host_err = PTR_ERR(dchild); - if (IS_ERR(dchild)) - return nfserrno(host_err); + if (IS_ERR(dchild)) { + err = nfserrno(host_err); + goto out_unlock; + } err = fh_compose(resfhp, fhp->fh_export, dchild, fhp); /* * We unconditionally drop our ref to dchild as fh_compose will have @@ -1337,175 +1404,14 @@ nfsd_create(struct svc_rqst *rqstp, struct svc_fh *fhp, */ dput(dchild); if (err) - return err; - return nfsd_create_locked(rqstp, fhp, fname, flen, iap, type, - rdev, resfhp); -} - -#ifdef CONFIG_NFSD_V3 - -/* - * NFSv3 and NFSv4 version of nfsd_create - */ -__be32 -do_nfsd_create(struct svc_rqst *rqstp, struct svc_fh *fhp, - char *fname, int flen, struct iattr *iap, - struct svc_fh *resfhp, int createmode, u32 *verifier, - bool *truncp, bool *created) -{ - struct dentry *dentry, *dchild = NULL; - struct inode *dirp; - __be32 err; - int host_err; - __u32 v_mtime=0, v_atime=0; - - err = nfserr_perm; - if (!flen) - goto out; - err = nfserr_exist; - if (isdotent(fname, flen)) - goto out; - if (!(iap->ia_valid & ATTR_MODE)) - iap->ia_mode = 0; - err = fh_verify(rqstp, fhp, S_IFDIR, NFSD_MAY_EXEC); - if (err) - goto out; - - dentry = fhp->fh_dentry; - dirp = d_inode(dentry); - - host_err = fh_want_write(fhp); - if (host_err) - goto out_nfserr; - - fh_lock_nested(fhp, I_MUTEX_PARENT); - - /* - * Compose the response file handle. - */ - dchild = lookup_one_len(fname, dentry, flen); - host_err = PTR_ERR(dchild); - if (IS_ERR(dchild)) - goto out_nfserr; - - /* If file doesn't exist, check for permissions to create one */ - if (d_really_is_negative(dchild)) { - err = fh_verify(rqstp, fhp, S_IFDIR, NFSD_MAY_CREATE); - if (err) - goto out; - } - - err = fh_compose(resfhp, fhp->fh_export, dchild, fhp); - if (err) - goto out; - - if (nfsd_create_is_exclusive(createmode)) { - /* solaris7 gets confused (bugid 4218508) if these have - * the high bit set, so just clear the high bits. If this is - * ever changed to use different attrs for storing the - * verifier, then do_open_lookup() will also need to be fixed - * accordingly. - */ - v_mtime = verifier[0]&0x7fffffff; - v_atime = verifier[1]&0x7fffffff; - } - - if (d_really_is_positive(dchild)) { - err = 0; - - switch (createmode) { - case NFS3_CREATE_UNCHECKED: - if (! d_is_reg(dchild)) - goto out; - else if (truncp) { - /* in nfsv4, we need to treat this case a little - * differently. we don't want to truncate the - * file now; this would be wrong if the OPEN - * fails for some other reason. furthermore, - * if the size is nonzero, we should ignore it - * according to spec! - */ - *truncp = (iap->ia_valid & ATTR_SIZE) && !iap->ia_size; - } - else { - iap->ia_valid &= ATTR_SIZE; - goto set_attr; - } - break; - case NFS3_CREATE_EXCLUSIVE: - if ( d_inode(dchild)->i_mtime.tv_sec == v_mtime - && d_inode(dchild)->i_atime.tv_sec == v_atime - && d_inode(dchild)->i_size == 0 ) { - if (created) - *created = true; - break; - } - /* fall through */ - case NFS4_CREATE_EXCLUSIVE4_1: - if ( d_inode(dchild)->i_mtime.tv_sec == v_mtime - && d_inode(dchild)->i_atime.tv_sec == v_atime - && d_inode(dchild)->i_size == 0 ) { - if (created) - *created = true; - goto set_attr; - } - /* fall through */ - case NFS3_CREATE_GUARDED: - err = nfserr_exist; - } - fh_drop_write(fhp); - goto out; - } - - host_err = vfs_create(dirp, dchild, iap->ia_mode, true); - if (host_err < 0) { - fh_drop_write(fhp); - goto out_nfserr; - } - if (created) - *created = true; - - nfsd_check_ignore_resizing(iap); - - if (nfsd_create_is_exclusive(createmode)) { - /* Cram the verifier into atime/mtime */ - iap->ia_valid = ATTR_MTIME|ATTR_ATIME - | ATTR_MTIME_SET|ATTR_ATIME_SET; - /* XXX someone who knows this better please fix it for nsec */ - iap->ia_mtime.tv_sec = v_mtime; - iap->ia_atime.tv_sec = v_atime; - iap->ia_mtime.tv_nsec = 0; - iap->ia_atime.tv_nsec = 0; - } - - set_attr: - err = nfsd_create_setattr(rqstp, resfhp, iap); - - /* - * nfsd_create_setattr already committed the child - * (and possibly also the parent). - */ - if (!err) - err = nfserrno(commit_metadata(fhp)); - - /* - * Update the filehandle to get the new inode info. - */ - if (!err) - err = fh_update(resfhp); - - out: - fh_unlock(fhp); - if (dchild && !IS_ERR(dchild)) - dput(dchild); - fh_drop_write(fhp); - return err; - - out_nfserr: - err = nfserrno(host_err); - goto out; + goto out_unlock; + fh_fill_pre_attrs(fhp); + err = nfsd_create_locked(rqstp, fhp, attrs, type, rdev, resfhp); + fh_fill_post_attrs(fhp); +out_unlock: + inode_unlock(dentry->d_inode); + return err; } -#endif /* CONFIG_NFSD_V3 */ /* * Read a symlink. On entry, *lenp must contain the maximum path length that @@ -1545,15 +1451,25 @@ nfsd_readlink(struct svc_rqst *rqstp, struct svc_fh *fhp, char *buf, int *lenp) return 0; } -/* - * Create a symlink and look up its inode +/** + * nfsd_symlink - Create a symlink and look up its inode + * @rqstp: RPC transaction being executed + * @fhp: NFS filehandle of parent directory + * @fname: filename of the new symlink + * @flen: length of @fname + * @path: content of the new symlink (NUL-terminated) + * @attrs: requested attributes of new object + * @resfhp: NFS filehandle of new object + * * N.B. After this call _both_ fhp and resfhp need an fh_put + * + * Returns nfs_ok on success, or an nfsstat in network byte order. */ __be32 nfsd_symlink(struct svc_rqst *rqstp, struct svc_fh *fhp, - char *fname, int flen, - char *path, - struct svc_fh *resfhp) + char *fname, int flen, + char *path, struct nfsd_attrs *attrs, + struct svc_fh *resfhp) { struct dentry *dentry, *dnew; __be32 err, cerr; @@ -1571,33 +1487,35 @@ nfsd_symlink(struct svc_rqst *rqstp, struct svc_fh *fhp, goto out; host_err = fh_want_write(fhp); - if (host_err) - goto out_nfserr; + if (host_err) { + err = nfserrno(host_err); + goto out; + } - fh_lock(fhp); dentry = fhp->fh_dentry; + inode_lock_nested(dentry->d_inode, I_MUTEX_PARENT); dnew = lookup_one_len(fname, dentry, flen); - host_err = PTR_ERR(dnew); - if (IS_ERR(dnew)) - goto out_nfserr; - - host_err = vfs_symlink(d_inode(dentry), dnew, path); + if (IS_ERR(dnew)) { + err = nfserrno(PTR_ERR(dnew)); + inode_unlock(dentry->d_inode); + goto out_drop_write; + } + fh_fill_pre_attrs(fhp); + host_err = vfs_symlink(&init_user_ns, d_inode(dentry), dnew, path); err = nfserrno(host_err); + cerr = fh_compose(resfhp, fhp->fh_export, dnew, fhp); + if (!err) + nfsd_create_setattr(rqstp, fhp, resfhp, attrs); + fh_fill_post_attrs(fhp); + inode_unlock(dentry->d_inode); if (!err) err = nfserrno(commit_metadata(fhp)); - fh_unlock(fhp); - - fh_drop_write(fhp); - - cerr = fh_compose(resfhp, fhp->fh_export, dnew, fhp); dput(dnew); if (err==0) err = cerr; +out_drop_write: + fh_drop_write(fhp); out: return err; - -out_nfserr: - err = nfserrno(host_err); - goto out; } /* @@ -1635,21 +1553,25 @@ nfsd_link(struct svc_rqst *rqstp, struct svc_fh *ffhp, goto out; } - fh_lock_nested(ffhp, I_MUTEX_PARENT); ddir = ffhp->fh_dentry; dirp = d_inode(ddir); + inode_lock_nested(dirp, I_MUTEX_PARENT); dnew = lookup_one_len(name, ddir, len); - host_err = PTR_ERR(dnew); - if (IS_ERR(dnew)) - goto out_nfserr; + if (IS_ERR(dnew)) { + err = nfserrno(PTR_ERR(dnew)); + goto out_unlock; + } dold = tfhp->fh_dentry; err = nfserr_noent; if (d_really_is_negative(dold)) goto out_dput; - host_err = vfs_link(dold, dirp, dnew, NULL); + fh_fill_pre_attrs(ffhp); + host_err = vfs_link(dold, &init_user_ns, dirp, dnew, NULL); + fh_fill_post_attrs(ffhp); + inode_unlock(dirp); if (!host_err) { err = nfserrno(commit_metadata(ffhp)); if (!err) @@ -1660,17 +1582,17 @@ nfsd_link(struct svc_rqst *rqstp, struct svc_fh *ffhp, else err = nfserrno(host_err); } -out_dput: dput(dnew); -out_unlock: - fh_unlock(ffhp); +out_drop_write: fh_drop_write(tfhp); out: return err; -out_nfserr: - err = nfserrno(host_err); - goto out_unlock; +out_dput: + dput(dnew); +out_unlock: + inode_unlock(dirp); + goto out_drop_write; } static void @@ -1705,7 +1627,7 @@ nfsd_rename(struct svc_rqst *rqstp, struct svc_fh *ffhp, char *fname, int flen, struct inode *fdir, *tdir; __be32 err; int host_err; - bool has_cached = false; + bool close_cached = false; err = fh_verify(rqstp, ffhp, S_IFDIR, NFSD_MAY_REMOVE); if (err) @@ -1731,12 +1653,9 @@ retry: goto out; } - /* cannot use fh_lock as we need deadlock protective ordering - * so do it by hand */ trap = lock_rename(tdentry, fdentry); - ffhp->fh_locked = tfhp->fh_locked = true; - fill_pre_wcc(ffhp); - fill_pre_wcc(tfhp); + fh_fill_pre_attrs(ffhp); + fh_fill_pre_attrs(tfhp); odentry = lookup_one_len(fname, fdentry, flen); host_err = PTR_ERR(odentry); @@ -1764,11 +1683,28 @@ retry: if (ffhp->fh_export->ex_path.dentry != tfhp->fh_export->ex_path.dentry) goto out_dput_new; - if (nfsd_has_cached_files(ndentry)) { - has_cached = true; + if ((ndentry->d_sb->s_export_op->flags & EXPORT_OP_CLOSE_BEFORE_UNLINK) && + nfsd_has_cached_files(ndentry)) { + close_cached = true; goto out_dput_old; } else { - host_err = vfs_rename(fdir, odentry, tdir, ndentry, NULL, 0); + struct renamedata rd = { + .old_mnt_userns = &init_user_ns, + .old_dir = fdir, + .old_dentry = odentry, + .new_mnt_userns = &init_user_ns, + .new_dir = tdir, + .new_dentry = ndentry, + }; + int retries; + + for (retries = 1;;) { + host_err = vfs_rename(&rd); + if (host_err != -EAGAIN || !retries--) + break; + if (!nfsd_wait_for_delegreturn(rqstp, d_inode(odentry))) + break; + } if (!host_err) { host_err = commit_metadata(tfhp); if (!host_err) @@ -1781,17 +1717,12 @@ retry: dput(odentry); out_nfserr: err = nfserrno(host_err); - /* - * We cannot rely on fh_unlock on the two filehandles, - * as that would do the wrong thing if the two directories - * were the same, so again we do it by hand. - */ - if (!has_cached) { - fill_post_wcc(ffhp); - fill_post_wcc(tfhp); + + if (!close_cached) { + fh_fill_post_attrs(ffhp); + fh_fill_post_attrs(tfhp); } unlock_rename(tdentry, fdentry); - ffhp->fh_locked = tfhp->fh_locked = false; fh_drop_write(ffhp); /* @@ -1800,8 +1731,8 @@ retry: * shouldn't be done with locks held however, so we delay it until this * point and then reattempt the whole shebang. */ - if (has_cached) { - has_cached = false; + if (close_cached) { + close_cached = false; nfsd_close_cached_files(ndentry); dput(ndentry); goto retry; @@ -1820,6 +1751,7 @@ nfsd_unlink(struct svc_rqst *rqstp, struct svc_fh *fhp, int type, { struct dentry *dentry, *rdentry; struct inode *dirp; + struct inode *rinode; __be32 err; int host_err; @@ -1834,34 +1766,50 @@ nfsd_unlink(struct svc_rqst *rqstp, struct svc_fh *fhp, int type, if (host_err) goto out_nfserr; - fh_lock_nested(fhp, I_MUTEX_PARENT); dentry = fhp->fh_dentry; dirp = d_inode(dentry); + inode_lock_nested(dirp, I_MUTEX_PARENT); rdentry = lookup_one_len(fname, dentry, flen); host_err = PTR_ERR(rdentry); if (IS_ERR(rdentry)) - goto out_drop_write; + goto out_unlock; if (d_really_is_negative(rdentry)) { dput(rdentry); host_err = -ENOENT; - goto out_drop_write; + goto out_unlock; } + rinode = d_inode(rdentry); + ihold(rinode); if (!type) type = d_inode(rdentry)->i_mode & S_IFMT; + fh_fill_pre_attrs(fhp); if (type != S_IFDIR) { - nfsd_close_cached_files(rdentry); - host_err = vfs_unlink(dirp, rdentry, NULL); + int retries; + + if (rdentry->d_sb->s_export_op->flags & EXPORT_OP_CLOSE_BEFORE_UNLINK) + nfsd_close_cached_files(rdentry); + + for (retries = 1;;) { + host_err = vfs_unlink(&init_user_ns, dirp, rdentry, NULL); + if (host_err != -EAGAIN || !retries--) + break; + if (!nfsd_wait_for_delegreturn(rqstp, rinode)) + break; + } } else { - host_err = vfs_rmdir(dirp, rdentry); + host_err = vfs_rmdir(&init_user_ns, dirp, rdentry); } + fh_fill_post_attrs(fhp); + inode_unlock(dirp); if (!host_err) host_err = commit_metadata(fhp); dput(rdentry); + iput(rinode); /* truncate the inode here */ out_drop_write: fh_drop_write(fhp); @@ -1879,6 +1827,9 @@ out_nfserr: } out: return err; +out_unlock: + inode_unlock(dirp); + goto out_drop_write; } /* @@ -1903,7 +1854,7 @@ struct readdir_data { int full; }; -static int nfsd_buffered_filldir(struct dir_context *ctx, const char *name, +static bool nfsd_buffered_filldir(struct dir_context *ctx, const char *name, int namlen, loff_t offset, u64 ino, unsigned int d_type) { @@ -1915,7 +1866,7 @@ static int nfsd_buffered_filldir(struct dir_context *ctx, const char *name, reclen = ALIGN(sizeof(struct buffered_dirent) + namlen, sizeof(u64)); if (buf->used + reclen > PAGE_SIZE) { buf->full = 1; - return -EINVAL; + return false; } de->namlen = namlen; @@ -1925,11 +1876,12 @@ static int nfsd_buffered_filldir(struct dir_context *ctx, const char *name, memcpy(de->name, name, namlen); buf->used += reclen; - return 0; + return true; } -static __be32 nfsd_buffered_readdir(struct file *file, nfsd_filldir_t func, - struct readdir_cd *cdp, loff_t *offsetp) +static __be32 nfsd_buffered_readdir(struct file *file, struct svc_fh *fhp, + nfsd_filldir_t func, struct readdir_cd *cdp, + loff_t *offsetp) { struct buffered_dirent *de; int host_err; @@ -1975,6 +1927,8 @@ static __be32 nfsd_buffered_readdir(struct file *file, nfsd_filldir_t func, if (cdp->err != nfs_ok) break; + trace_nfsd_dirent(fhp, de->ino, de->name, de->namlen); + reclen = ALIGN(sizeof(*de) + de->namlen, sizeof(u64)); size -= reclen; @@ -2022,7 +1976,7 @@ nfsd_readdir(struct svc_rqst *rqstp, struct svc_fh *fhp, loff_t *offsetp, goto out_close; } - err = nfsd_buffered_readdir(file, func, cdp, offsetp); + err = nfsd_buffered_readdir(file, fhp, func, cdp, offsetp); if (err == nfserr_eof || err == nfserr_toosmall) err = nfs_ok; /* can still be found in ->err */ @@ -2058,6 +2012,242 @@ static int exp_rdonly(struct svc_rqst *rqstp, struct svc_export *exp) return nfsexp_flags(rqstp, exp) & NFSEXP_READONLY; } +#ifdef CONFIG_NFSD_V4 +/* + * Helper function to translate error numbers. In the case of xattr operations, + * some error codes need to be translated outside of the standard translations. + * + * ENODATA needs to be translated to nfserr_noxattr. + * E2BIG to nfserr_xattr2big. + * + * Additionally, vfs_listxattr can return -ERANGE. This means that the + * file has too many extended attributes to retrieve inside an + * XATTR_LIST_MAX sized buffer. This is a bug in the xattr implementation: + * filesystems will allow the adding of extended attributes until they hit + * their own internal limit. This limit may be larger than XATTR_LIST_MAX. + * So, at that point, the attributes are present and valid, but can't + * be retrieved using listxattr, since the upper level xattr code enforces + * the XATTR_LIST_MAX limit. + * + * This bug means that we need to deal with listxattr returning -ERANGE. The + * best mapping is to return TOOSMALL. + */ +static __be32 +nfsd_xattr_errno(int err) +{ + switch (err) { + case -ENODATA: + return nfserr_noxattr; + case -E2BIG: + return nfserr_xattr2big; + case -ERANGE: + return nfserr_toosmall; + } + return nfserrno(err); +} + +/* + * Retrieve the specified user extended attribute. To avoid always + * having to allocate the maximum size (since we are not getting + * a maximum size from the RPC), do a probe + alloc. Hold a reader + * lock on i_rwsem to prevent the extended attribute from changing + * size while we're doing this. + */ +__be32 +nfsd_getxattr(struct svc_rqst *rqstp, struct svc_fh *fhp, char *name, + void **bufp, int *lenp) +{ + ssize_t len; + __be32 err; + char *buf; + struct inode *inode; + struct dentry *dentry; + + err = fh_verify(rqstp, fhp, 0, NFSD_MAY_READ); + if (err) + return err; + + err = nfs_ok; + dentry = fhp->fh_dentry; + inode = d_inode(dentry); + + inode_lock_shared(inode); + + len = vfs_getxattr(&init_user_ns, dentry, name, NULL, 0); + + /* + * Zero-length attribute, just return. + */ + if (len == 0) { + *bufp = NULL; + *lenp = 0; + goto out; + } + + if (len < 0) { + err = nfsd_xattr_errno(len); + goto out; + } + + if (len > *lenp) { + err = nfserr_toosmall; + goto out; + } + + buf = kvmalloc(len, GFP_KERNEL | GFP_NOFS); + if (buf == NULL) { + err = nfserr_jukebox; + goto out; + } + + len = vfs_getxattr(&init_user_ns, dentry, name, buf, len); + if (len <= 0) { + kvfree(buf); + buf = NULL; + err = nfsd_xattr_errno(len); + } + + *lenp = len; + *bufp = buf; + +out: + inode_unlock_shared(inode); + + return err; +} + +/* + * Retrieve the xattr names. Since we can't know how many are + * user extended attributes, we must get all attributes here, + * and have the XDR encode filter out the "user." ones. + * + * While this could always just allocate an XATTR_LIST_MAX + * buffer, that's a waste, so do a probe + allocate. To + * avoid any changes between the probe and allocate, wrap + * this in inode_lock. + */ +__be32 +nfsd_listxattr(struct svc_rqst *rqstp, struct svc_fh *fhp, char **bufp, + int *lenp) +{ + ssize_t len; + __be32 err; + char *buf; + struct inode *inode; + struct dentry *dentry; + + err = fh_verify(rqstp, fhp, 0, NFSD_MAY_READ); + if (err) + return err; + + dentry = fhp->fh_dentry; + inode = d_inode(dentry); + *lenp = 0; + + inode_lock_shared(inode); + + len = vfs_listxattr(dentry, NULL, 0); + if (len <= 0) { + err = nfsd_xattr_errno(len); + goto out; + } + + if (len > XATTR_LIST_MAX) { + err = nfserr_xattr2big; + goto out; + } + + /* + * We're holding i_rwsem - use GFP_NOFS. + */ + buf = kvmalloc(len, GFP_KERNEL | GFP_NOFS); + if (buf == NULL) { + err = nfserr_jukebox; + goto out; + } + + len = vfs_listxattr(dentry, buf, len); + if (len <= 0) { + kvfree(buf); + err = nfsd_xattr_errno(len); + goto out; + } + + *lenp = len; + *bufp = buf; + + err = nfs_ok; +out: + inode_unlock_shared(inode); + + return err; +} + +/** + * nfsd_removexattr - Remove an extended attribute + * @rqstp: RPC transaction being executed + * @fhp: NFS filehandle of object with xattr to remove + * @name: name of xattr to remove (NUL-terminate) + * + * Pass in a NULL pointer for delegated_inode, and let the client deal + * with NFS4ERR_DELAY (same as with e.g. setattr and remove). + * + * Returns nfs_ok on success, or an nfsstat in network byte order. + */ +__be32 +nfsd_removexattr(struct svc_rqst *rqstp, struct svc_fh *fhp, char *name) +{ + __be32 err; + int ret; + + err = fh_verify(rqstp, fhp, 0, NFSD_MAY_WRITE); + if (err) + return err; + + ret = fh_want_write(fhp); + if (ret) + return nfserrno(ret); + + inode_lock(fhp->fh_dentry->d_inode); + fh_fill_pre_attrs(fhp); + + ret = __vfs_removexattr_locked(&init_user_ns, fhp->fh_dentry, + name, NULL); + + fh_fill_post_attrs(fhp); + inode_unlock(fhp->fh_dentry->d_inode); + fh_drop_write(fhp); + + return nfsd_xattr_errno(ret); +} + +__be32 +nfsd_setxattr(struct svc_rqst *rqstp, struct svc_fh *fhp, char *name, + void *buf, u32 len, u32 flags) +{ + __be32 err; + int ret; + + err = fh_verify(rqstp, fhp, 0, NFSD_MAY_WRITE); + if (err) + return err; + + ret = fh_want_write(fhp); + if (ret) + return nfserrno(ret); + inode_lock(fhp->fh_dentry->d_inode); + fh_fill_pre_attrs(fhp); + + ret = __vfs_setxattr_locked(&init_user_ns, fhp->fh_dentry, name, buf, + len, flags, NULL); + fh_fill_post_attrs(fhp); + inode_unlock(fhp->fh_dentry->d_inode); + fh_drop_write(fhp); + + return nfsd_xattr_errno(ret); +} +#endif + /* * Check for a user's access permissions to this inode. */ @@ -2132,13 +2322,14 @@ nfsd_permission(struct svc_rqst *rqstp, struct svc_export *exp, return 0; /* This assumes NFSD_MAY_{READ,WRITE,EXEC} == MAY_{READ,WRITE,EXEC} */ - err = inode_permission(inode, acc & (MAY_READ|MAY_WRITE|MAY_EXEC)); + err = inode_permission(&init_user_ns, inode, + acc & (MAY_READ | MAY_WRITE | MAY_EXEC)); /* Allow read access to binaries even when mode 111 */ if (err == -EACCES && S_ISREG(inode->i_mode) && (acc == (NFSD_MAY_READ | NFSD_MAY_OWNER_OVERRIDE) || acc == (NFSD_MAY_READ | NFSD_MAY_READ_IF_EXEC))) - err = inode_permission(inode, MAY_EXEC); + err = inode_permission(&init_user_ns, inode, MAY_EXEC); return err? nfserrno(err) : 0; } |