Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs-2.6v2.6.33-rc4-git6

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs-2.6:
  do_add_mount() should sanitize mnt_flags
  CIFS shouldn't make mountpoints shrinkable
  mnt_flags fixes in do_remount()
  attach_recursive_mnt() needs to hold vfsmount_lock over set_mnt_shared()
  may_umount() needs namespace_sem
  Fix configfs leak
  Fix the -ESTALE handling in do_filp_open()
  ecryptfs: Fix refcnt leak on ecryptfs_follow_link() error path
  Fix ACC_MODE() for real
  Unrot uml mconsole a bit
  hppfs: handle ->put_link()
  Kill 9p readlink()
  fix autofs/afs/etc. magic mountpoint breakage

11 files changed, 71 insertions, 115 deletions

diff --git a/arch/um/drivers/mconsole_kern.c b/arch/um/drivers/mconsole_kern.c
index 51069245b79a..3b3c36601a7b 100644
--- a/arch/um/drivers/mconsole_kern.c
+++ b/arch/um/drivers/mconsole_kern.c
@@ -125,50 +125,36 @@ void mconsole_log(struct mc_request *req)
 void mconsole_proc(struct mc_request *req)
 {
 	struct nameidata nd;
-	struct file_system_type *proc;
-	struct super_block *super;
+	struct vfsmount *mnt = current->nsproxy->pid_ns->proc_mnt;
 	struct file *file;
 	int n, err;
 	char *ptr = req->request.data, *buf;
+	mm_segment_t old_fs = get_fs();
 
 	ptr += strlen("proc");
 	ptr = skip_spaces(ptr);
 
-	proc = get_fs_type("proc");
-	if (proc == NULL) {
-		mconsole_reply(req, "procfs not registered", 1, 0);
+	err = vfs_path_lookup(mnt->mnt_root, mnt, ptr, LOOKUP_FOLLOW, &nd);
+	if (err) {
+		mconsole_reply(req, "Failed to look up file", 1, 0);
 		goto out;
 	}
 
-	super = (*proc->get_sb)(proc, 0, NULL, NULL);
-	put_filesystem(proc);
-	if (super == NULL) {
-		mconsole_reply(req, "Failed to get procfs superblock", 1, 0);
+	err = may_open(&nd.path, MAY_READ, FMODE_READ);
+	if (result) {
+		mconsole_reply(req, "Failed to open file", 1, 0);
+		path_put(&nd.path);
 		goto out;
 	}
-	up_write(&super->s_umount);
-
-	nd.path.dentry = super->s_root;
-	nd.path.mnt = NULL;
-	nd.flags = O_RDONLY + 1;
-	nd.last_type = LAST_ROOT;
-
-	/* START: it was experienced that the stability problems are closed
-	 * if commenting out these two calls + the below read cycle. To
-	 * make UML crash again, it was enough to readd either one.*/
-	err = link_path_walk(ptr, &nd);
-	if (err) {
-		mconsole_reply(req, "Failed to look up file", 1, 0);
-		goto out_kill;
-	}
 
 	file = dentry_open(nd.path.dentry, nd.path.mnt, O_RDONLY,
 			   current_cred());
+	err = PTR_ERR(file);
 	if (IS_ERR(file)) {
 		mconsole_reply(req, "Failed to open file", 1, 0);
-		goto out_kill;
+		path_put(&nd.path);
+		goto out;
 	}
-	/*END*/
 
 	buf = kmalloc(PAGE_SIZE, GFP_KERNEL);
 	if (buf == NULL) {
@@ -176,10 +162,13 @@ void mconsole_proc(struct mc_request *req)
 		goto out_fput;
 	}
 
-	if ((file->f_op != NULL) && (file->f_op->read != NULL)) {
+	if (file->f_op->read) {
 		do {
-			n = (*file->f_op->read)(file, buf, PAGE_SIZE - 1,
-						&file->f_pos);
+			loff_t pos;
+			set_fs(KERNEL_DS);
+			n = vfs_read(file, buf, PAGE_SIZE - 1, &pos);
+			file_pos_write(file, pos);
+			set_fs(old_fs);
 			if (n >= 0) {
 				buf[n] = '\0';
 				mconsole_reply(req, buf, 0, (n > 0));
@@ -197,8 +186,6 @@ void mconsole_proc(struct mc_request *req)
 	kfree(buf);
  out_fput:
 	fput(file);
- out_kill:
-	deactivate_super(super);
  out: ;
 }
 #endif
diff --git a/fs/9p/vfs_inode.c b/fs/9p/vfs_inode.c
index 18f74ec4dce9..9d03d1ebca6f 100644
--- a/fs/9p/vfs_inode.c
+++ b/fs/9p/vfs_inode.c
@@ -1001,44 +1001,6 @@ done:
 }
 
 /**
- * v9fs_vfs_readlink - read a symlink's location
- * @dentry: dentry for symlink
- * @buffer: buffer to load symlink location into
- * @buflen: length of buffer
- *
- */
-
-static int v9fs_vfs_readlink(struct dentry *dentry, char __user * buffer,
-			     int buflen)
-{
-	int retval;
-	int ret;
-	char *link = __getname();
-
-	if (unlikely(!link))
-		return -ENOMEM;
-
-	if (buflen > PATH_MAX)
-		buflen = PATH_MAX;
-
-	P9_DPRINTK(P9_DEBUG_VFS, " dentry: %s (%p)\n", dentry->d_name.name,
-									dentry);
-
-	retval = v9fs_readlink(dentry, link, buflen);
-
-	if (retval > 0) {
-		if ((ret = copy_to_user(buffer, link, retval)) != 0) {
-			P9_DPRINTK(P9_DEBUG_ERROR,
-					"problem copying to user: %d\n", ret);
-			retval = ret;
-		}
-	}
-
-	__putname(link);
-	return retval;
-}
-
-/**
  * v9fs_vfs_follow_link - follow a symlink path
  * @dentry: dentry for symlink
  * @nd: nameidata
@@ -1230,7 +1192,6 @@ static const struct inode_operations v9fs_dir_inode_operations_ext = {
 	.rmdir = v9fs_vfs_rmdir,
 	.mknod = v9fs_vfs_mknod,
 	.rename = v9fs_vfs_rename,
-	.readlink = v9fs_vfs_readlink,
 	.getattr = v9fs_vfs_getattr,
 	.setattr = v9fs_vfs_setattr,
 };
@@ -1253,7 +1214,7 @@ static const struct inode_operations v9fs_file_inode_operations = {
 };
 
 static const struct inode_operations v9fs_symlink_inode_operations = {
-	.readlink = v9fs_vfs_readlink,
+	.readlink = generic_readlink,
 	.follow_link = v9fs_vfs_follow_link,
 	.put_link = v9fs_vfs_put_link,
 	.getattr = v9fs_vfs_getattr,
diff --git a/fs/cifs/cifs_dfs_ref.c b/fs/cifs/cifs_dfs_ref.c
index fea9e898c4ba..b44ce0a0711c 100644
--- a/fs/cifs/cifs_dfs_ref.c
+++ b/fs/cifs/cifs_dfs_ref.c
@@ -269,7 +269,7 @@ static int add_mount_helper(struct vfsmount *newmnt, struct nameidata *nd,
 	int err;
 
 	mntget(newmnt);
-	err = do_add_mount(newmnt, &nd->path, nd->path.mnt->mnt_flags, mntlist);
+	err = do_add_mount(newmnt, &nd->path, nd->path.mnt->mnt_flags | MNT_SHRINKABLE, mntlist);
 	switch (err) {
 	case 0:
 		path_put(&nd->path);
@@ -371,7 +371,6 @@ cifs_dfs_follow_mountpoint(struct dentry *dentry, struct nameidata *nd)
 	if (IS_ERR(mnt))
 		goto out_err;
 
-	nd->path.mnt->mnt_flags |= MNT_SHRINKABLE;
 	rc = add_mount_helper(mnt, nd, &cifs_dfs_automount_list);
 
 out:
diff --git a/fs/configfs/symlink.c b/fs/configfs/symlink.c
index c8afa6b1d91d..32a5f46b1157 100644
--- a/fs/configfs/symlink.c
+++ b/fs/configfs/symlink.c
@@ -121,8 +121,10 @@ static int get_target(const char *symname, struct path *path,
 				ret = -ENOENT;
 				path_put(path);
 			}
-		} else
+		} else {
 			ret = -EPERM;
+			path_put(path);
+		}
 	}
 
 	return ret;
diff --git a/fs/ecryptfs/inode.c b/fs/ecryptfs/inode.c
index 429ca0b3ba08..7f8545032930 100644
--- a/fs/ecryptfs/inode.c
+++ b/fs/ecryptfs/inode.c
@@ -715,31 +715,31 @@ static void *ecryptfs_follow_link(struct dentry *dentry, struct nameidata *nd)
 	/* Released in ecryptfs_put_link(); only release here on error */
 	buf = kmalloc(len, GFP_KERNEL);
 	if (!buf) {
-		rc = -ENOMEM;
+		buf = ERR_PTR(-ENOMEM);
 		goto out;
 	}
 	old_fs = get_fs();
 	set_fs(get_ds());
 	rc = dentry->d_inode->i_op->readlink(dentry, (char __user *)buf, len);
 	set_fs(old_fs);
-	if (rc < 0)
-		goto out_free;
-	else
+	if (rc < 0) {
+		kfree(buf);
+		buf = ERR_PTR(rc);
+	} else
 		buf[rc] = '\0';
-	rc = 0;
-	nd_set_link(nd, buf);
-	goto out;
-out_free:
-	kfree(buf);
 out:
-	return ERR_PTR(rc);
+	nd_set_link(nd, buf);
+	return NULL;
 }
 
 static void
 ecryptfs_put_link(struct dentry *dentry, struct nameidata *nd, void *ptr)
 {
-	/* Free the char* */
-	kfree(nd_get_link(nd));
+	char *buf = nd_get_link(nd);
+	if (!IS_ERR(buf)) {
+		/* Free the char* */
+		kfree(buf);
+	}
 }
 
 /**
diff --git a/fs/hppfs/hppfs.c b/fs/hppfs/hppfs.c
index a5089a6dd67a..7239efc690d8 100644
--- a/fs/hppfs/hppfs.c
+++ b/fs/hppfs/hppfs.c
@@ -646,22 +646,27 @@ static const struct super_operations hppfs_sbops = {
 static int hppfs_readlink(struct dentry *dentry, char __user *buffer,
 			  int buflen)
 {
-	struct dentry *proc_dentry;
-
-	proc_dentry = HPPFS_I(dentry->d_inode)->proc_dentry;
+	struct dentry *proc_dentry = HPPFS_I(dentry->d_inode)->proc_dentry;
 	return proc_dentry->d_inode->i_op->readlink(proc_dentry, buffer,
 						    buflen);
 }
 
 static void *hppfs_follow_link(struct dentry *dentry, struct nameidata *nd)
 {
-	struct dentry *proc_dentry;
-
-	proc_dentry = HPPFS_I(dentry->d_inode)->proc_dentry;
+	struct dentry *proc_dentry = HPPFS_I(dentry->d_inode)->proc_dentry;
 
 	return proc_dentry->d_inode->i_op->follow_link(proc_dentry, nd);
 }
 
+static void hppfs_put_link(struct dentry *dentry, struct nameidata *nd,
+			   void *cookie)
+{
+	struct dentry *proc_dentry = HPPFS_I(dentry->d_inode)->proc_dentry;
+
+	if (proc_dentry->d_inode->i_op->put_link)
+		proc_dentry->d_inode->i_op->put_link(proc_dentry, nd, cookie);
+}
+
 static const struct inode_operations hppfs_dir_iops = {
 	.lookup		= hppfs_lookup,
 };
@@ -669,6 +674,7 @@ static const struct inode_operations hppfs_dir_iops = {
 static const struct inode_operations hppfs_link_iops = {
 	.readlink	= hppfs_readlink,
 	.follow_link	= hppfs_follow_link,
+	.put_link	= hppfs_put_link,
 };
 
 static struct inode *get_inode(struct super_block *sb, struct dentry *dentry)
diff --git a/fs/namei.c b/fs/namei.c
index b55440baf7ab..94a5e60779f9 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -561,6 +561,7 @@ static __always_inline int __do_follow_link(struct path *path, struct nameidata
 		dget(dentry);
 	}
 	mntget(path->mnt);
+	nd->last_type = LAST_BIND;
 	cookie = dentry->d_inode->i_op->follow_link(dentry, nd);
 	error = PTR_ERR(cookie);
 	if (!IS_ERR(cookie)) {
@@ -1603,11 +1604,12 @@ struct file *do_filp_open(int dfd, const char *pathname,
 	struct file *filp;
 	struct nameidata nd;
 	int error;
-	struct path path, save;
+	struct path path;
 	struct dentry *dir;
 	int count = 0;
 	int will_truncate;
 	int flag = open_to_namei_flags(open_flag);
+	int force_reval = 0;
 
 	/*
 	 * O_SYNC is implemented as __O_SYNC|O_DSYNC.  As many places only
@@ -1619,7 +1621,7 @@ struct file *do_filp_open(int dfd, const char *pathname,
 		open_flag |= O_DSYNC;
 
 	if (!acc_mode)
-		acc_mode = MAY_OPEN | ACC_MODE(flag);
+		acc_mode = MAY_OPEN | ACC_MODE(open_flag);
 
 	/* O_TRUNC implies we need access checks for write permissions */
 	if (flag & O_TRUNC)
@@ -1659,9 +1661,12 @@ struct file *do_filp_open(int dfd, const char *pathname,
 	/*
 	 * Create - we need to know the parent.
 	 */
+reval:
 	error = path_init(dfd, pathname, LOOKUP_PARENT, &nd);
 	if (error)
 		return ERR_PTR(error);
+	if (force_reval)
+		nd.flags |= LOOKUP_REVAL;
 	error = path_walk(pathname, &nd);
 	if (error) {
 		if (nd.root.mnt)
@@ -1853,17 +1858,7 @@ do_link:
 	error = security_inode_follow_link(path.dentry, &nd);
 	if (error)
 		goto exit_dput;
-	save = nd.path;
-	path_get(&save);
 	error = __do_follow_link(&path, &nd);
-	if (error == -ESTALE) {
-		/* nd.path had been dropped */
-		nd.path = save;
-		path_get(&nd.path);
-		nd.flags |= LOOKUP_REVAL;
-		error = __do_follow_link(&path, &nd);
-	}
-	path_put(&save);
 	path_put(&path);
 	if (error) {
 		/* Does someone understand code flow here? Or it is only
@@ -1873,6 +1868,10 @@ do_link:
 		release_open_intent(&nd);
 		if (nd.root.mnt)
 			path_put(&nd.root);
+		if (error == -ESTALE && !force_reval) {
+			force_reval = 1;
+			goto reval;
+		}
 		return ERR_PTR(error);
 	}
 	nd.flags &= ~LOOKUP_PARENT;
diff --git a/fs/namespace.c b/fs/namespace.c
index 7d70d63ceb29..c768f733c8d6 100644
--- a/fs/namespace.c
+++ b/fs/namespace.c
@@ -965,10 +965,12 @@ EXPORT_SYMBOL(may_umount_tree);
 int may_umount(struct vfsmount *mnt)
 {
 	int ret = 1;
+	down_read(&namespace_sem);
 	spin_lock(&vfsmount_lock);
 	if (propagate_mount_busy(mnt, 2))
 		ret = 0;
 	spin_unlock(&vfsmount_lock);
+	up_read(&namespace_sem);
 	return ret;
 }
 
@@ -1352,12 +1354,12 @@ static int attach_recursive_mnt(struct vfsmount *source_mnt,
 	if (err)
 		goto out_cleanup_ids;
 
+	spin_lock(&vfsmount_lock);
+
 	if (IS_MNT_SHARED(dest_mnt)) {
 		for (p = source_mnt; p; p = next_mnt(p, source_mnt))
 			set_mnt_shared(p);
 	}
-
-	spin_lock(&vfsmount_lock);
 	if (parent_path) {
 		detach_mnt(source_mnt, parent_path);
 		attach_mnt(source_mnt, path);
@@ -1534,8 +1536,12 @@ static int do_remount(struct path *path, int flags, int mnt_flags,
 		err = change_mount_flags(path->mnt, flags);
 	else
 		err = do_remount_sb(sb, flags, data, 0);
-	if (!err)
+	if (!err) {
+		spin_lock(&vfsmount_lock);
+		mnt_flags |= path->mnt->mnt_flags & MNT_PNODE_MASK;
 		path->mnt->mnt_flags = mnt_flags;
+		spin_unlock(&vfsmount_lock);
+	}
 	up_write(&sb->s_umount);
 	if (!err) {
 		security_sb_post_remount(path->mnt, flags, data);
@@ -1665,6 +1671,8 @@ int do_add_mount(struct vfsmount *newmnt, struct path *path,
 {
 	int err;
 
+	mnt_flags &= ~(MNT_SHARED | MNT_WRITE_HOLD);
+
 	down_write(&namespace_sem);
 	/* Something was mounted here while we slept */
 	while (d_mountpoint(path->dentry) &&
diff --git a/fs/proc/base.c b/fs/proc/base.c
index 18d5cc62d8ed..e42bbd843ed1 100644
--- a/fs/proc/base.c
+++ b/fs/proc/base.c
@@ -1419,7 +1419,6 @@ static void *proc_pid_follow_link(struct dentry *dentry, struct nameidata *nd)
 		goto out;
 
 	error = PROC_I(inode)->op.proc_get_link(inode, &nd->path);
-	nd->last_type = LAST_BIND;
 out:
 	return ERR_PTR(error);
 }
diff --git a/include/linux/fs.h b/include/linux/fs.h
index 9147ca88f253..b1bcb275b596 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -2463,7 +2463,7 @@ int proc_nr_files(struct ctl_table *table, int write,
 
 int __init get_filesystem_list(char *buf);
 
-#define ACC_MODE(x) ("\000\004\002\006"[(x)&O_ACCMODE])
+#define ACC_MODE(x) ("\004\002\006\006"[(x)&O_ACCMODE])
 #define OPEN_FMODE(flag) ((__force fmode_t)((flag + 1) & O_ACCMODE))
 
 #endif /* __KERNEL__ */
diff --git a/security/tomoyo/tomoyo.c b/security/tomoyo/tomoyo.c
index 8a00ade85166..2aceebf5f354 100644
--- a/security/tomoyo/tomoyo.c
+++ b/security/tomoyo/tomoyo.c
@@ -80,9 +80,8 @@ static int tomoyo_bprm_check_security(struct linux_binprm *bprm)
 		return tomoyo_find_next_domain(bprm);
 	/*
 	 * Read permission is checked against interpreters using next domain.
-	 * '1' is the result of open_to_namei_flags(O_RDONLY).
 	 */
-	return tomoyo_check_open_permission(domain, &bprm->file->f_path, 1);
+	return tomoyo_check_open_permission(domain, &bprm->file->f_path, O_RDONLY);
 }
 
 static int tomoyo_path_truncate(struct path *path, loff_t length,
@@ -184,10 +183,6 @@ static int tomoyo_file_fcntl(struct file *file, unsigned int cmd,
 static int tomoyo_dentry_open(struct file *f, const struct cred *cred)
 {
 	int flags = f->f_flags;
-
-	if ((flags + 1) & O_ACCMODE)
-		flags++;
-	flags |= f->f_flags & (O_APPEND | O_TRUNC);
 	/* Don't check read permission here if called from do_execve(). */
 	if (current->in_execve)
 		return 0;