Commit 1171afc4 authored by Eric W. Biederman's avatar Eric W. Biederman Committed by Sasha Levin

mnt: Add a per mount namespace limit on the number of mounts

[ Upstream commit d2921684 ]

CAI Qian <caiqian@redhat.com> pointed out that the semantics
of shared subtrees make it possible to create an exponentially
increasing number of mounts in a mount namespace.

    mkdir /tmp/1 /tmp/2
    mount --make-rshared /
    for i in $(seq 1 20) ; do mount --bind /tmp/1 /tmp/2 ; done

Will create create 2^20 or 1048576 mounts, which is a practical problem
as some people have managed to hit this by accident.

As such CVE-2016-6213 was assigned.

Ian Kent <raven@themaw.net> described the situation for autofs users
as follows:

> The number of mounts for direct mount maps is usually not very large because of
> the way they are implemented, large direct mount maps can have performance
> problems. There can be anywhere from a few (likely case a few hundred) to less
> than 10000, plus mounts that have been triggered and not yet expired.
>
> Indirect mounts have one autofs mount at the root plus the number of mounts that
> have been triggered and not yet expired.
>
> The number of autofs indirect map entries can range from a few to the common
> case of several thousand and in rare cases up to between 30000 and 50000. I've
> not heard of people with maps larger than 50000 entries.
>
> The larger the number of map entries the greater the possibility for a large
> number of active mounts so it's not hard to expect cases of a 1000 or somewhat
> more active mounts.

So I am setting the default number of mounts allowed per mount
namespace at 100,000.  This is more than enough for any use case I
know of, but small enough to quickly stop an exponential increase
in mounts.  Which should be perfect to catch misconfigurations and
malfunctioning programs.

For anyone who needs a higher limit this can be changed by writing
to the new /proc/sys/fs/mount-max sysctl.
Tested-by: default avatarCAI Qian <caiqian@redhat.com>
Signed-off-by: default avatar"Eric W. Biederman" <ebiederm@xmission.com>

Conflicts:
	fs/namespace.c
	kernel/sysctl.c
Signed-off-by: default avatarPhilipp Hahn <hahn@univention.de>
Signed-off-by: default avatarSasha Levin <alexander.levin@verizon.com>
parent 62fa696b
...@@ -265,6 +265,13 @@ aio-nr can grow to. ...@@ -265,6 +265,13 @@ aio-nr can grow to.
============================================================== ==============================================================
mount-max:
This denotes the maximum number of mounts that may exist
in a mount namespace.
==============================================================
2. /proc/sys/fs/binfmt_misc 2. /proc/sys/fs/binfmt_misc
---------------------------------------------------------- ----------------------------------------------------------
......
...@@ -13,6 +13,8 @@ struct mnt_namespace { ...@@ -13,6 +13,8 @@ struct mnt_namespace {
u64 seq; /* Sequence number to prevent loops */ u64 seq; /* Sequence number to prevent loops */
wait_queue_head_t poll; wait_queue_head_t poll;
u64 event; u64 event;
unsigned int mounts; /* # of mounts in the namespace */
unsigned int pending_mounts;
}; };
struct mnt_pcp { struct mnt_pcp {
......
...@@ -27,6 +27,9 @@ ...@@ -27,6 +27,9 @@
#include "pnode.h" #include "pnode.h"
#include "internal.h" #include "internal.h"
/* Maximum number of mounts in a mount namespace */
unsigned int sysctl_mount_max __read_mostly = 100000;
static unsigned int m_hash_mask __read_mostly; static unsigned int m_hash_mask __read_mostly;
static unsigned int m_hash_shift __read_mostly; static unsigned int m_hash_shift __read_mostly;
static unsigned int mp_hash_mask __read_mostly; static unsigned int mp_hash_mask __read_mostly;
...@@ -888,6 +891,9 @@ static void commit_tree(struct mount *mnt, struct mount *shadows) ...@@ -888,6 +891,9 @@ static void commit_tree(struct mount *mnt, struct mount *shadows)
list_splice(&head, n->list.prev); list_splice(&head, n->list.prev);
n->mounts += n->pending_mounts;
n->pending_mounts = 0;
attach_shadowed(mnt, parent, shadows); attach_shadowed(mnt, parent, shadows);
touch_mnt_namespace(n); touch_mnt_namespace(n);
} }
...@@ -1408,11 +1414,16 @@ static void umount_tree(struct mount *mnt, enum umount_tree_flags how) ...@@ -1408,11 +1414,16 @@ static void umount_tree(struct mount *mnt, enum umount_tree_flags how)
propagate_umount(&tmp_list); propagate_umount(&tmp_list);
while (!list_empty(&tmp_list)) { while (!list_empty(&tmp_list)) {
struct mnt_namespace *ns;
bool disconnect; bool disconnect;
p = list_first_entry(&tmp_list, struct mount, mnt_list); p = list_first_entry(&tmp_list, struct mount, mnt_list);
list_del_init(&p->mnt_expire); list_del_init(&p->mnt_expire);
list_del_init(&p->mnt_list); list_del_init(&p->mnt_list);
__touch_mnt_namespace(p->mnt_ns); ns = p->mnt_ns;
if (ns) {
ns->mounts--;
__touch_mnt_namespace(ns);
}
p->mnt_ns = NULL; p->mnt_ns = NULL;
if (how & UMOUNT_SYNC) if (how & UMOUNT_SYNC)
p->mnt.mnt_flags |= MNT_SYNC_UMOUNT; p->mnt.mnt_flags |= MNT_SYNC_UMOUNT;
...@@ -1821,6 +1832,28 @@ static int invent_group_ids(struct mount *mnt, bool recurse) ...@@ -1821,6 +1832,28 @@ static int invent_group_ids(struct mount *mnt, bool recurse)
return 0; return 0;
} }
int count_mounts(struct mnt_namespace *ns, struct mount *mnt)
{
unsigned int max = READ_ONCE(sysctl_mount_max);
unsigned int mounts = 0, old, pending, sum;
struct mount *p;
for (p = mnt; p; p = next_mnt(p, mnt))
mounts++;
old = ns->mounts;
pending = ns->pending_mounts;
sum = old + pending;
if ((old > sum) ||
(pending > sum) ||
(max < sum) ||
(mounts > (max - sum)))
return -ENOSPC;
ns->pending_mounts = pending + mounts;
return 0;
}
/* /*
* @source_mnt : mount tree to be attached * @source_mnt : mount tree to be attached
* @nd : place the mount tree @source_mnt is attached * @nd : place the mount tree @source_mnt is attached
...@@ -1890,10 +1923,18 @@ static int attach_recursive_mnt(struct mount *source_mnt, ...@@ -1890,10 +1923,18 @@ static int attach_recursive_mnt(struct mount *source_mnt,
struct path *parent_path) struct path *parent_path)
{ {
HLIST_HEAD(tree_list); HLIST_HEAD(tree_list);
struct mnt_namespace *ns = dest_mnt->mnt_ns;
struct mount *child, *p; struct mount *child, *p;
struct hlist_node *n; struct hlist_node *n;
int err; int err;
/* Is there space to add these mounts to the mount namespace? */
if (!parent_path) {
err = count_mounts(ns, source_mnt);
if (err)
goto out;
}
if (IS_MNT_SHARED(dest_mnt)) { if (IS_MNT_SHARED(dest_mnt)) {
err = invent_group_ids(source_mnt, true); err = invent_group_ids(source_mnt, true);
if (err) if (err)
...@@ -1930,11 +1971,13 @@ static int attach_recursive_mnt(struct mount *source_mnt, ...@@ -1930,11 +1971,13 @@ static int attach_recursive_mnt(struct mount *source_mnt,
out_cleanup_ids: out_cleanup_ids:
while (!hlist_empty(&tree_list)) { while (!hlist_empty(&tree_list)) {
child = hlist_entry(tree_list.first, struct mount, mnt_hash); child = hlist_entry(tree_list.first, struct mount, mnt_hash);
child->mnt_parent->mnt_ns->pending_mounts = 0;
umount_tree(child, UMOUNT_SYNC); umount_tree(child, UMOUNT_SYNC);
} }
unlock_mount_hash(); unlock_mount_hash();
cleanup_group_ids(source_mnt, NULL); cleanup_group_ids(source_mnt, NULL);
out: out:
ns->pending_mounts = 0;
return err; return err;
} }
...@@ -2758,6 +2801,8 @@ static struct mnt_namespace *alloc_mnt_ns(struct user_namespace *user_ns) ...@@ -2758,6 +2801,8 @@ static struct mnt_namespace *alloc_mnt_ns(struct user_namespace *user_ns)
init_waitqueue_head(&new_ns->poll); init_waitqueue_head(&new_ns->poll);
new_ns->event = 0; new_ns->event = 0;
new_ns->user_ns = get_user_ns(user_ns); new_ns->user_ns = get_user_ns(user_ns);
new_ns->mounts = 0;
new_ns->pending_mounts = 0;
return new_ns; return new_ns;
} }
...@@ -2807,6 +2852,7 @@ struct mnt_namespace *copy_mnt_ns(unsigned long flags, struct mnt_namespace *ns, ...@@ -2807,6 +2852,7 @@ struct mnt_namespace *copy_mnt_ns(unsigned long flags, struct mnt_namespace *ns,
q = new; q = new;
while (p) { while (p) {
q->mnt_ns = new_ns; q->mnt_ns = new_ns;
new_ns->mounts++;
if (new_fs) { if (new_fs) {
if (&p->mnt == new_fs->root.mnt) { if (&p->mnt == new_fs->root.mnt) {
new_fs->root.mnt = mntget(&q->mnt); new_fs->root.mnt = mntget(&q->mnt);
...@@ -2845,6 +2891,7 @@ static struct mnt_namespace *create_mnt_ns(struct vfsmount *m) ...@@ -2845,6 +2891,7 @@ static struct mnt_namespace *create_mnt_ns(struct vfsmount *m)
struct mount *mnt = real_mount(m); struct mount *mnt = real_mount(m);
mnt->mnt_ns = new_ns; mnt->mnt_ns = new_ns;
new_ns->root = mnt; new_ns->root = mnt;
new_ns->mounts++;
list_add(&mnt->mnt_list, &new_ns->list); list_add(&mnt->mnt_list, &new_ns->list);
} else { } else {
mntput(m); mntput(m);
......
...@@ -259,7 +259,7 @@ static int propagate_one(struct mount *m) ...@@ -259,7 +259,7 @@ static int propagate_one(struct mount *m)
read_sequnlock_excl(&mount_lock); read_sequnlock_excl(&mount_lock);
} }
hlist_add_head(&child->mnt_hash, list); hlist_add_head(&child->mnt_hash, list);
return 0; return count_mounts(m->mnt_ns, child);
} }
/* /*
......
...@@ -52,4 +52,5 @@ void mnt_set_mountpoint(struct mount *, struct mountpoint *, ...@@ -52,4 +52,5 @@ void mnt_set_mountpoint(struct mount *, struct mountpoint *,
struct mount *copy_tree(struct mount *, struct dentry *, int); struct mount *copy_tree(struct mount *, struct dentry *, int);
bool is_path_reachable(struct mount *, struct dentry *, bool is_path_reachable(struct mount *, struct dentry *,
const struct path *root); const struct path *root);
int count_mounts(struct mnt_namespace *ns, struct mount *mnt);
#endif /* _LINUX_PNODE_H */ #endif /* _LINUX_PNODE_H */
...@@ -95,4 +95,6 @@ extern void mark_mounts_for_expiry(struct list_head *mounts); ...@@ -95,4 +95,6 @@ extern void mark_mounts_for_expiry(struct list_head *mounts);
extern dev_t name_to_dev_t(const char *name); extern dev_t name_to_dev_t(const char *name);
extern unsigned int sysctl_mount_max;
#endif /* _LINUX_MOUNT_H */ #endif /* _LINUX_MOUNT_H */
...@@ -64,6 +64,7 @@ ...@@ -64,6 +64,7 @@
#include <linux/binfmts.h> #include <linux/binfmts.h>
#include <linux/sched/sysctl.h> #include <linux/sched/sysctl.h>
#include <linux/kexec.h> #include <linux/kexec.h>
#include <linux/mount.h>
#include <asm/uaccess.h> #include <asm/uaccess.h>
#include <asm/processor.h> #include <asm/processor.h>
...@@ -1709,6 +1710,14 @@ static struct ctl_table fs_table[] = { ...@@ -1709,6 +1710,14 @@ static struct ctl_table fs_table[] = {
.mode = 0644, .mode = 0644,
.proc_handler = proc_doulongvec_minmax, .proc_handler = proc_doulongvec_minmax,
}, },
{
.procname = "mount-max",
.data = &sysctl_mount_max,
.maxlen = sizeof(unsigned int),
.mode = 0644,
.proc_handler = proc_dointvec_minmax,
.extra1 = &one,
},
{ } { }
}; };
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment