1 files changed, 221 insertions, 38 deletions

diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c
index 80ba3950c40d..ff15522481d4 100644
--- a/fs/fuse/dir.c
+++ b/fs/fuse/dir.c
@@ -14,6 +14,29 @@
 #include <linux/namei.h>
 #include <linux/slab.h>
 
+static bool fuse_use_readdirplus(struct inode *dir, struct file *filp)
+{
+	struct fuse_conn *fc = get_fuse_conn(dir);
+	struct fuse_inode *fi = get_fuse_inode(dir);
+
+	if (!fc->do_readdirplus)
+		return false;
+	if (!fc->readdirplus_auto)
+		return true;
+	if (test_and_clear_bit(FUSE_I_ADVISE_RDPLUS, &fi->state))
+		return true;
+	if (filp->f_pos == 0)
+		return true;
+	return false;
+}
+
+static void fuse_advise_use_readdirplus(struct inode *dir)
+{
+	struct fuse_inode *fi = get_fuse_inode(dir);
+
+	set_bit(FUSE_I_ADVISE_RDPLUS, &fi->state);
+}
+
 #if BITS_PER_LONG >= 64
 static inline void fuse_dentry_settime(struct dentry *entry, u64 time)
 {
@@ -178,7 +201,7 @@ static int fuse_dentry_revalidate(struct dentry *entry, unsigned int flags)
 			return -ECHILD;
 
 		fc = get_fuse_conn(inode);
-		req = fuse_get_req(fc);
+		req = fuse_get_req_nopages(fc);
 		if (IS_ERR(req))
 			return 0;
 
@@ -219,6 +242,7 @@ static int fuse_dentry_revalidate(struct dentry *entry, unsigned int flags)
 				       attr_version);
 		fuse_change_entry_timeout(entry, &outarg);
 	}
+	fuse_advise_use_readdirplus(inode);
 	return 1;
 }
 
@@ -271,7 +295,7 @@ int fuse_lookup_name(struct super_block *sb, u64 nodeid, struct qstr *name,
 	if (name->len > FUSE_NAME_MAX)
 		goto out;
 
-	req = fuse_get_req(fc);
+	req = fuse_get_req_nopages(fc);
 	err = PTR_ERR(req);
 	if (IS_ERR(req))
 		goto out;
@@ -355,6 +379,7 @@ static struct dentry *fuse_lookup(struct inode *dir, struct dentry *entry,
 	else
 		fuse_invalidate_entry_cache(entry);
 
+	fuse_advise_use_readdirplus(dir);
 	return newent;
 
  out_iput:
@@ -391,7 +416,7 @@ static int fuse_create_open(struct inode *dir, struct dentry *entry,
 	if (!forget)
 		goto out_err;
 
-	req = fuse_get_req(fc);
+	req = fuse_get_req_nopages(fc);
 	err = PTR_ERR(req);
 	if (IS_ERR(req))
 		goto out_put_forget_req;
@@ -592,7 +617,7 @@ static int fuse_mknod(struct inode *dir, struct dentry *entry, umode_t mode,
 {
 	struct fuse_mknod_in inarg;
 	struct fuse_conn *fc = get_fuse_conn(dir);
-	struct fuse_req *req = fuse_get_req(fc);
+	struct fuse_req *req = fuse_get_req_nopages(fc);
 	if (IS_ERR(req))
 		return PTR_ERR(req);
 
@@ -623,7 +648,7 @@ static int fuse_mkdir(struct inode *dir, struct dentry *entry, umode_t mode)
 {
 	struct fuse_mkdir_in inarg;
 	struct fuse_conn *fc = get_fuse_conn(dir);
-	struct fuse_req *req = fuse_get_req(fc);
+	struct fuse_req *req = fuse_get_req_nopages(fc);
 	if (IS_ERR(req))
 		return PTR_ERR(req);
 
@@ -647,7 +672,7 @@ static int fuse_symlink(struct inode *dir, struct dentry *entry,
 {
 	struct fuse_conn *fc = get_fuse_conn(dir);
 	unsigned len = strlen(link) + 1;
-	struct fuse_req *req = fuse_get_req(fc);
+	struct fuse_req *req = fuse_get_req_nopages(fc);
 	if (IS_ERR(req))
 		return PTR_ERR(req);
 
@@ -664,7 +689,7 @@ static int fuse_unlink(struct inode *dir, struct dentry *entry)
 {
 	int err;
 	struct fuse_conn *fc = get_fuse_conn(dir);
-	struct fuse_req *req = fuse_get_req(fc);
+	struct fuse_req *req = fuse_get_req_nopages(fc);
 	if (IS_ERR(req))
 		return PTR_ERR(req);
 
@@ -682,7 +707,14 @@ static int fuse_unlink(struct inode *dir, struct dentry *entry)
 
 		spin_lock(&fc->lock);
 		fi->attr_version = ++fc->attr_version;
-		drop_nlink(inode);
+		/*
+		 * If i_nlink == 0 then unlink doesn't make sense, yet this can
+		 * happen if userspace filesystem is careless.  It would be
+		 * difficult to enforce correct nlink usage so just ignore this
+		 * condition here
+		 */
+		if (inode->i_nlink > 0)
+			drop_nlink(inode);
 		spin_unlock(&fc->lock);
 		fuse_invalidate_attr(inode);
 		fuse_invalidate_attr(dir);
@@ -696,7 +728,7 @@ static int fuse_rmdir(struct inode *dir, struct dentry *entry)
 {
 	int err;
 	struct fuse_conn *fc = get_fuse_conn(dir);
-	struct fuse_req *req = fuse_get_req(fc);
+	struct fuse_req *req = fuse_get_req_nopages(fc);
 	if (IS_ERR(req))
 		return PTR_ERR(req);
 
@@ -723,7 +755,7 @@ static int fuse_rename(struct inode *olddir, struct dentry *oldent,
 	int err;
 	struct fuse_rename_in inarg;
 	struct fuse_conn *fc = get_fuse_conn(olddir);
-	struct fuse_req *req = fuse_get_req(fc);
+	struct fuse_req *req = fuse_get_req_nopages(fc);
 
 	if (IS_ERR(req))
 		return PTR_ERR(req);
@@ -776,7 +808,7 @@ static int fuse_link(struct dentry *entry, struct inode *newdir,
 	struct fuse_link_in inarg;
 	struct inode *inode = entry->d_inode;
 	struct fuse_conn *fc = get_fuse_conn(inode);
-	struct fuse_req *req = fuse_get_req(fc);
+	struct fuse_req *req = fuse_get_req_nopages(fc);
 	if (IS_ERR(req))
 		return PTR_ERR(req);
 
@@ -848,7 +880,7 @@ static int fuse_do_getattr(struct inode *inode, struct kstat *stat,
 	struct fuse_req *req;
 	u64 attr_version;
 
-	req = fuse_get_req(fc);
+	req = fuse_get_req_nopages(fc);
 	if (IS_ERR(req))
 		return PTR_ERR(req);
 
@@ -985,7 +1017,7 @@ int fuse_reverse_inval_entry(struct super_block *sb, u64 parent_nodeid,
 
 /*
  * Calling into a user-controlled filesystem gives the filesystem
- * daemon ptrace-like capabilities over the requester process.  This
+ * daemon ptrace-like capabilities over the current process.  This
  * means, that the filesystem daemon is able to record the exact
  * filesystem operations performed, and can also control the behavior
  * of the requester process in otherwise impossible ways.  For example
@@ -996,27 +1028,23 @@ int fuse_reverse_inval_entry(struct super_block *sb, u64 parent_nodeid,
  * for which the owner of the mount has ptrace privilege.  This
  * excludes processes started by other users, suid or sgid processes.
  */
-int fuse_allow_task(struct fuse_conn *fc, struct task_struct *task)
+int fuse_allow_current_process(struct fuse_conn *fc)
 {
 	const struct cred *cred;
-	int ret;
 
 	if (fc->flags & FUSE_ALLOW_OTHER)
 		return 1;
 
-	rcu_read_lock();
-	ret = 0;
-	cred = __task_cred(task);
+	cred = current_cred();
 	if (uid_eq(cred->euid, fc->user_id) &&
 	    uid_eq(cred->suid, fc->user_id) &&
 	    uid_eq(cred->uid,  fc->user_id) &&
 	    gid_eq(cred->egid, fc->group_id) &&
 	    gid_eq(cred->sgid, fc->group_id) &&
 	    gid_eq(cred->gid,  fc->group_id))
-		ret = 1;
-	rcu_read_unlock();
+		return 1;
 
-	return ret;
+	return 0;
 }
 
 static int fuse_access(struct inode *inode, int mask)
@@ -1029,7 +1057,7 @@ static int fuse_access(struct inode *inode, int mask)
 	if (fc->no_access)
 		return 0;
 
-	req = fuse_get_req(fc);
+	req = fuse_get_req_nopages(fc);
 	if (IS_ERR(req))
 		return PTR_ERR(req);
 
@@ -1077,7 +1105,7 @@ static int fuse_permission(struct inode *inode, int mask)
 	bool refreshed = false;
 	int err = 0;
 
-	if (!fuse_allow_task(fc, current))
+	if (!fuse_allow_current_process(fc))
 		return -EACCES;
 
 	/*
@@ -1155,19 +1183,157 @@ static int parse_dirfile(char *buf, size_t nbytes, struct file *file,
 	return 0;
 }
 
-static int fuse_readdir(struct file *file, void *dstbuf, filldir_t filldir)
+static int fuse_direntplus_link(struct file *file,
+				struct fuse_direntplus *direntplus,
+				u64 attr_version)
 {
 	int err;
+	struct fuse_entry_out *o = &direntplus->entry_out;
+	struct fuse_dirent *dirent = &direntplus->dirent;
+	struct dentry *parent = file->f_path.dentry;
+	struct qstr name = QSTR_INIT(dirent->name, dirent->namelen);
+	struct dentry *dentry;
+	struct dentry *alias;
+	struct inode *dir = parent->d_inode;
+	struct fuse_conn *fc;
+	struct inode *inode;
+
+	if (!o->nodeid) {
+		/*
+		 * Unlike in the case of fuse_lookup, zero nodeid does not mean
+		 * ENOENT. Instead, it only means the userspace filesystem did
+		 * not want to return attributes/handle for this entry.
+		 *
+		 * So do nothing.
+		 */
+		return 0;
+	}
+
+	if (name.name[0] == '.') {
+		/*
+		 * We could potentially refresh the attributes of the directory
+		 * and its parent?
+		 */
+		if (name.len == 1)
+			return 0;
+		if (name.name[1] == '.' && name.len == 2)
+			return 0;
+	}
+	fc = get_fuse_conn(dir);
+
+	name.hash = full_name_hash(name.name, name.len);
+	dentry = d_lookup(parent, &name);
+	if (dentry && dentry->d_inode) {
+		inode = dentry->d_inode;
+		if (get_node_id(inode) == o->nodeid) {
+			struct fuse_inode *fi;
+			fi = get_fuse_inode(inode);
+			spin_lock(&fc->lock);
+			fi->nlookup++;
+			spin_unlock(&fc->lock);
+
+			/*
+			 * The other branch to 'found' comes via fuse_iget()
+			 * which bumps nlookup inside
+			 */
+			goto found;
+		}
+		err = d_invalidate(dentry);
+		if (err)
+			goto out;
+		dput(dentry);
+		dentry = NULL;
+	}
+
+	dentry = d_alloc(parent, &name);
+	err = -ENOMEM;
+	if (!dentry)
+		goto out;
+
+	inode = fuse_iget(dir->i_sb, o->nodeid, o->generation,
+			  &o->attr, entry_attr_timeout(o), attr_version);
+	if (!inode)
+		goto out;
+
+	alias = d_materialise_unique(dentry, inode);
+	err = PTR_ERR(alias);
+	if (IS_ERR(alias))
+		goto out;
+	if (alias) {
+		dput(dentry);
+		dentry = alias;
+	}
+
+found:
+	fuse_change_attributes(inode, &o->attr, entry_attr_timeout(o),
+			       attr_version);
+
+	fuse_change_entry_timeout(dentry, o);
+
+	err = 0;
+out:
+	if (dentry)
+		dput(dentry);
+	return err;
+}
+
+static int parse_dirplusfile(char *buf, size_t nbytes, struct file *file,
+			     void *dstbuf, filldir_t filldir, u64 attr_version)
+{
+	struct fuse_direntplus *direntplus;
+	struct fuse_dirent *dirent;
+	size_t reclen;
+	int over = 0;
+	int ret;
+
+	while (nbytes >= FUSE_NAME_OFFSET_DIRENTPLUS) {
+		direntplus = (struct fuse_direntplus *) buf;
+		dirent = &direntplus->dirent;
+		reclen = FUSE_DIRENTPLUS_SIZE(direntplus);
+
+		if (!dirent->namelen || dirent->namelen > FUSE_NAME_MAX)
+			return -EIO;
+		if (reclen > nbytes)
+			break;
+
+		if (!over) {
+			/* We fill entries into dstbuf only as much as
+			   it can hold. But we still continue iterating
+			   over remaining entries to link them. If not,
+			   we need to send a FORGET for each of those
+			   which we did not link.
+			*/
+			over = filldir(dstbuf, dirent->name, dirent->namelen,
+				       file->f_pos, dirent->ino,
+				       dirent->type);
+			file->f_pos = dirent->off;
+		}
+
+		buf += reclen;
+		nbytes -= reclen;
+
+		ret = fuse_direntplus_link(file, direntplus, attr_version);
+		if (ret)
+			fuse_force_forget(file, direntplus->entry_out.nodeid);
+	}
+
+	return 0;
+}
+
+static int fuse_readdir(struct file *file, void *dstbuf, filldir_t filldir)
+{
+	int plus, err;
 	size_t nbytes;
 	struct page *page;
 	struct inode *inode = file_inode(file);
 	struct fuse_conn *fc = get_fuse_conn(inode);
 	struct fuse_req *req;
+	u64 attr_version = 0;
 
 	if (is_bad_inode(inode))
 		return -EIO;
 
-	req = fuse_get_req(fc);
+	req = fuse_get_req(fc, 1);
 	if (IS_ERR(req))
 		return PTR_ERR(req);
 
@@ -1176,17 +1342,34 @@ static int fuse_readdir(struct file *file, void *dstbuf, filldir_t filldir)
 		fuse_put_request(fc, req);
 		return -ENOMEM;
 	}
+
+	plus = fuse_use_readdirplus(inode, file);
 	req->out.argpages = 1;
 	req->num_pages = 1;
 	req->pages[0] = page;
-	fuse_read_fill(req, file, file->f_pos, PAGE_SIZE, FUSE_READDIR);
+	req->page_descs[0].length = PAGE_SIZE;
+	if (plus) {
+		attr_version = fuse_get_attr_version(fc);
+		fuse_read_fill(req, file, file->f_pos, PAGE_SIZE,
+			       FUSE_READDIRPLUS);
+	} else {
+		fuse_read_fill(req, file, file->f_pos, PAGE_SIZE,
+			       FUSE_READDIR);
+	}
 	fuse_request_send(fc, req);
 	nbytes = req->out.args[0].size;
 	err = req->out.h.error;
 	fuse_put_request(fc, req);
-	if (!err)
-		err = parse_dirfile(page_address(page), nbytes, file, dstbuf,
-				    filldir);
+	if (!err) {
+		if (plus) {
+			err = parse_dirplusfile(page_address(page), nbytes,
+						file, dstbuf, filldir,
+						attr_version);
+		} else {
+			err = parse_dirfile(page_address(page), nbytes, file,
+					    dstbuf, filldir);
+		}
+	}
 
 	__free_page(page);
 	fuse_invalidate_attr(inode); /* atime changed */
@@ -1197,7 +1380,7 @@ static char *read_link(struct dentry *dentry)
 {
 	struct inode *inode = dentry->d_inode;
 	struct fuse_conn *fc = get_fuse_conn(inode);
-	struct fuse_req *req = fuse_get_req(fc);
+	struct fuse_req *req = fuse_get_req_nopages(fc);
 	char *link;
 
 	if (IS_ERR(req))
@@ -1391,7 +1574,7 @@ static int fuse_do_setattr(struct dentry *entry, struct iattr *attr,
 	loff_t oldsize;
 	int err;
 
-	if (!fuse_allow_task(fc, current))
+	if (!fuse_allow_current_process(fc))
 		return -EACCES;
 
 	if (!(fc->flags & FUSE_DEFAULT_PERMISSIONS))
@@ -1410,7 +1593,7 @@ static int fuse_do_setattr(struct dentry *entry, struct iattr *attr,
 	if (attr->ia_valid & ATTR_SIZE)
 		is_truncate = true;
 
-	req = fuse_get_req(fc);
+	req = fuse_get_req_nopages(fc);
 	if (IS_ERR(req))
 		return PTR_ERR(req);
 
@@ -1500,7 +1683,7 @@ static int fuse_getattr(struct vfsmount *mnt, struct dentry *entry,
 	struct inode *inode = entry->d_inode;
 	struct fuse_conn *fc = get_fuse_conn(inode);
 
-	if (!fuse_allow_task(fc, current))
+	if (!fuse_allow_current_process(fc))
 		return -EACCES;
 
 	return fuse_update_attributes(inode, stat, NULL, NULL);
@@ -1518,7 +1701,7 @@ static int fuse_setxattr(struct dentry *entry, const char *name,
 	if (fc->no_setxattr)
 		return -EOPNOTSUPP;
 
-	req = fuse_get_req(fc);
+	req = fuse_get_req_nopages(fc);
 	if (IS_ERR(req))
 		return PTR_ERR(req);
 
@@ -1557,7 +1740,7 @@ static ssize_t fuse_getxattr(struct dentry *entry, const char *name,
 	if (fc->no_getxattr)
 		return -EOPNOTSUPP;
 
-	req = fuse_get_req(fc);
+	req = fuse_get_req_nopages(fc);
 	if (IS_ERR(req))
 		return PTR_ERR(req);
 
@@ -1603,13 +1786,13 @@ static ssize_t fuse_listxattr(struct dentry *entry, char *list, size_t size)
 	struct fuse_getxattr_out outarg;
 	ssize_t ret;
 
-	if (!fuse_allow_task(fc, current))
+	if (!fuse_allow_current_process(fc))
 		return -EACCES;
 
 	if (fc->no_listxattr)
 		return -EOPNOTSUPP;
 
-	req = fuse_get_req(fc);
+	req = fuse_get_req_nopages(fc);
 	if (IS_ERR(req))
 		return PTR_ERR(req);
 
@@ -1654,7 +1837,7 @@ static int fuse_removexattr(struct dentry *entry, const char *name)
 	if (fc->no_removexattr)
 		return -EOPNOTSUPP;
 
-	req = fuse_get_req(fc);
+	req = fuse_get_req_nopages(fc);
 	if (IS_ERR(req))
 		return PTR_ERR(req);