linux-kernel - [RFC PATCH 7/8] autofs - use path_has_submounts() to fix unreliable have

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20161003004659.4865.60429.stgit@pluto.themaw.net>
Date:   Mon, 03 Oct 2016 08:46:59 +0800
From:   Ian Kent <raven@...maw.net>
To:     Kernel Mailing List <linux-kernel@...r.kernel.org>
Cc:     autofs mailing list <autofs@...r.kernel.org>,
        Al Viro <viro@...IV.linux.org.uk>,
        linux-fsdevel <linux-fsdevel@...r.kernel.org>,
        Omar Sandoval <osandov@...ndov.com>,
        Andrew Morton <akpm@...ux-foundation.org>,
        "Eric W. Biederman" <ebiederm@...ssion.com>
Subject: [RFC PATCH 7/8] autofs - use path_has_submounts() to fix unreliable
 have_submount() checks

From: Ian Kent <ikent@...hat.com>

If an automount mount is clone(2)ed into a file system that is propagation
private, when it later expires in the originating namespace, subsequent
calls to autofs ->d_automount() for that dentry in the original namespace
will return ELOOP until the mount is umounted in the cloned namespace.

Now that a struct path is available where needed use path_has_submounts()
instead of have_submounts() so we don't get false positives when checking
if a dentry is a mount point or contains mounts in the current namespace.

Signed-off-by: Ian Kent <raven@...maw.net>
Cc: Al Viro <viro@...IV.linux.org.uk>
Cc: Eric W. Biederman <ebiederm@...ssion.com>
Cc: Omar Sandoval <osandov@...ndov.com>
---
 fs/autofs4/dev-ioctl.c |    2 +-
 fs/autofs4/root.c      |   14 +++++++-------
 fs/autofs4/waitq.c     |   10 +++++++---
 3 files changed, 15 insertions(+), 11 deletions(-)

diff --git a/fs/autofs4/dev-ioctl.c b/fs/autofs4/dev-ioctl.c
index 3cd96e6..953e790 100644
--- a/fs/autofs4/dev-ioctl.c
+++ b/fs/autofs4/dev-ioctl.c
@@ -564,7 +564,7 @@ static int autofs_dev_ioctl_ismountpoint(struct file *fp,
 
 		devid = new_encode_dev(dev);
 
-		err = have_submounts(path.dentry);
+		err = path_has_submounts(&path);
 
 		if (follow_down_one(&path))
 			magic = path.dentry->d_sb->s_magic;
diff --git a/fs/autofs4/root.c b/fs/autofs4/root.c
index 0f5d264..cf90d37 100644
--- a/fs/autofs4/root.c
+++ b/fs/autofs4/root.c
@@ -387,16 +387,16 @@ static struct vfsmount *autofs4_d_automount(struct path *path)
 		/*
 		 * It's possible that user space hasn't removed directories
 		 * after umounting a rootless multi-mount, although it
-		 * should. For v5 have_submounts() is sufficient to handle
-		 * this because the leaves of the directory tree under the
-		 * mount never trigger mounts themselves (they have an autofs
-		 * trigger mount mounted on them). But v4 pseudo direct mounts
-		 * do need the leaves to trigger mounts. In this case we
-		 * have no choice but to use the list_empty() check and
+		 * should. For v5 path_has_submounts() is sufficient to
+		 * handle this because the leaves of the directory tree under
+		 * the mount never trigger mounts themselves (they have an
+		 * autofs trigger mount mounted on them). But v4 pseudo direct
+		 * mounts do need the leaves to trigger mounts. In this case
+		 * we have no choice but to use the list_empty() check and
 		 * require user space behave.
 		 */
 		if (sbi->version > 4) {
-			if (have_submounts(dentry)) {
+			if (path_has_submounts(path)) {
 				spin_unlock(&sbi->fs_lock);
 				goto done;
 			}
diff --git a/fs/autofs4/waitq.c b/fs/autofs4/waitq.c
index f757f87..ed05cae 100644
--- a/fs/autofs4/waitq.c
+++ b/fs/autofs4/waitq.c
@@ -250,8 +250,9 @@ autofs4_find_wait(struct autofs_sb_info *sbi, const struct qstr *qstr)
 static int validate_request(struct autofs_wait_queue **wait,
 			    struct autofs_sb_info *sbi,
 			    const struct qstr *qstr,
-			    struct dentry *dentry, enum autofs_notify notify)
+			    struct path *path, enum autofs_notify notify)
 {
+	struct dentry *dentry = path->dentry;
 	struct autofs_wait_queue *wq;
 	struct autofs_info *ino;
 
@@ -314,6 +315,7 @@ static int validate_request(struct autofs_wait_queue **wait,
 	 */
 	if (notify == NFY_MOUNT) {
 		struct dentry *new = NULL;
+		struct path this;
 		int valid = 1;
 
 		/*
@@ -333,7 +335,9 @@ static int validate_request(struct autofs_wait_queue **wait,
 					dentry = new;
 			}
 		}
-		if (have_submounts(dentry))
+		this.mnt = path->mnt;
+		this.dentry = dentry;
+		if (path_has_submounts(&this))
 			valid = 0;
 
 		if (new)
@@ -406,7 +410,7 @@ int autofs4_wait(struct autofs_sb_info *sbi,
 		return -EINTR;
 	}
 
-	ret = validate_request(&wq, sbi, &qstr, dentry, notify);
+	ret = validate_request(&wq, sbi, &qstr, path, notify);
 	if (ret <= 0) {
 		if (ret != -EINTR)
 			mutex_unlock(&sbi->wq_mutex);