From: "\"Eric W. Beiderman" <ebiederm@xmission.com>
To: <linux-kernel@vger.kernel.org>
Cc: <linux-fsdevel@vger.kernel.org>,
<linux-security-module@vger.kernel.org>,
Linux Containers <containers@lists.linux-foundation.org>,
Andrew Morton <akpm@linux-foundation.org>,
Linus Torvalds <torvalds@linux-foundation.org>,
Al Viro <viro@ZenIV.linux.org.uk>,
Cyrill Gorcunov <gorcunov@openvz.org>,
"Eric W. Biederman" <ebiederm@xmission.com>
Subject: [PATCH 16/43] userns: Simplify the user_namespace by making userns->creator a kuid.
Date: Sat, 7 Apr 2012 22:15:12 -0700 [thread overview]
Message-ID: <1333862139-31737-16-git-send-email-ebiederm@xmission.com> (raw)
In-Reply-To: <m11unyn70b.fsf@fess.ebiederm.org>
From: Eric W. Biederman <ebiederm@xmission.com>
- Transform userns->creator from a user_struct reference to a simple
kuid_t, kgid_t pair.
In cap_capable this allows the check to see if we are the creator of
a namespace to become the classic suser style euid permission check.
This allows us to remove the need for a struct cred in the mapping
functions and still be able to dispaly the user namespace creators
uid and gid as 0.
- Remove the now unnecessary delayed_work in free_user_ns.
All that is left for free_user_ns to do is to call kmem_cache_free
and put_user_ns. Those functions can be called in any context
so call them directly from free_user_ns removing the need for delayed work.
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
---
include/linux/user_namespace.h | 4 ++--
kernel/user.c | 7 ++++---
kernel/user_namespace.c | 39 ++++++++++++++++++---------------------
security/commoncap.c | 5 +++--
4 files changed, 27 insertions(+), 28 deletions(-)
diff --git a/include/linux/user_namespace.h b/include/linux/user_namespace.h
index d767508..8a391bd 100644
--- a/include/linux/user_namespace.h
+++ b/include/linux/user_namespace.h
@@ -9,8 +9,8 @@
struct user_namespace {
struct kref kref;
struct user_namespace *parent;
- struct user_struct *creator;
- struct work_struct destroyer;
+ kuid_t owner;
+ kgid_t group;
};
extern struct user_namespace init_user_ns;
diff --git a/kernel/user.c b/kernel/user.c
index 025077e..cff3856 100644
--- a/kernel/user.c
+++ b/kernel/user.c
@@ -25,7 +25,8 @@ struct user_namespace init_user_ns = {
.kref = {
.refcount = ATOMIC_INIT(3),
},
- .creator = &root_user,
+ .owner = GLOBAL_ROOT_UID,
+ .group = GLOBAL_ROOT_GID,
};
EXPORT_SYMBOL_GPL(init_user_ns);
@@ -54,9 +55,9 @@ struct hlist_head uidhash_table[UIDHASH_SZ];
*/
static DEFINE_SPINLOCK(uidhash_lock);
-/* root_user.__count is 2, 1 for init task cred, 1 for init_user_ns->user_ns */
+/* root_user.__count is 1, for init task cred */
struct user_struct root_user = {
- .__count = ATOMIC_INIT(2),
+ .__count = ATOMIC_INIT(1),
.processes = ATOMIC_INIT(1),
.files = ATOMIC_INIT(0),
.sigpending = ATOMIC_INIT(0),
diff --git a/kernel/user_namespace.c b/kernel/user_namespace.c
index 898e973..f69741a 100644
--- a/kernel/user_namespace.c
+++ b/kernel/user_namespace.c
@@ -27,6 +27,16 @@ int create_user_ns(struct cred *new)
{
struct user_namespace *ns, *parent_ns = new->user_ns;
struct user_struct *root_user;
+ kuid_t owner = make_kuid(new->user_ns, new->euid);
+ kgid_t group = make_kgid(new->user_ns, new->egid);
+
+ /* The creator needs a mapping in the parent user namespace
+ * or else we won't be able to reasonably tell userspace who
+ * created a user_namespace.
+ */
+ if (!kuid_has_mapping(parent_ns, owner) ||
+ !kgid_has_mapping(parent_ns, group))
+ return -EPERM;
ns = kmem_cache_alloc(user_ns_cachep, GFP_KERNEL);
if (!ns)
@@ -43,7 +53,9 @@ int create_user_ns(struct cred *new)
/* set the new root user in the credentials under preparation */
ns->parent = parent_ns;
- ns->creator = new->user;
+ ns->owner = owner;
+ ns->group = group;
+ free_uid(new->user);
new->user = root_user;
new->uid = new->euid = new->suid = new->fsuid = 0;
new->gid = new->egid = new->sgid = new->fsgid = 0;
@@ -69,29 +81,15 @@ int create_user_ns(struct cred *new)
return 0;
}
-/*
- * Deferred destructor for a user namespace. This is required because
- * free_user_ns() may be called with uidhash_lock held, but we need to call
- * back to free_uid() which will want to take the lock again.
- */
-static void free_user_ns_work(struct work_struct *work)
+void free_user_ns(struct kref *kref)
{
struct user_namespace *parent, *ns =
- container_of(work, struct user_namespace, destroyer);
+ container_of(kref, struct user_namespace, kref);
+
parent = ns->parent;
- free_uid(ns->creator);
kmem_cache_free(user_ns_cachep, ns);
put_user_ns(parent);
}
-
-void free_user_ns(struct kref *kref)
-{
- struct user_namespace *ns =
- container_of(kref, struct user_namespace, kref);
-
- INIT_WORK(&ns->destroyer, free_user_ns_work);
- schedule_work(&ns->destroyer);
-}
EXPORT_SYMBOL(free_user_ns);
uid_t user_ns_map_uid(struct user_namespace *to, const struct cred *cred, uid_t uid)
@@ -101,12 +99,11 @@ uid_t user_ns_map_uid(struct user_namespace *to, const struct cred *cred, uid_t
if (likely(to == cred->user_ns))
return uid;
-
/* Is cred->user the creator of the target user_ns
* or the creator of one of it's parents?
*/
for ( tmp = to; tmp != &init_user_ns; tmp = tmp->parent ) {
- if (cred->user == tmp->creator) {
+ if (uid_eq(cred->user->uid, tmp->owner)) {
return (uid_t)0;
}
}
@@ -126,7 +123,7 @@ gid_t user_ns_map_gid(struct user_namespace *to, const struct cred *cred, gid_t
* or the creator of one of it's parents?
*/
for ( tmp = to; tmp != &init_user_ns; tmp = tmp->parent ) {
- if (cred->user == tmp->creator) {
+ if (uid_eq(cred->user->uid, tmp->owner)) {
return (gid_t)0;
}
}
diff --git a/security/commoncap.c b/security/commoncap.c
index 435d074..f2399d8 100644
--- a/security/commoncap.c
+++ b/security/commoncap.c
@@ -76,8 +76,9 @@ int cap_capable(const struct cred *cred, struct user_namespace *targ_ns,
int cap, int audit)
{
for (;;) {
- /* The creator of the user namespace has all caps. */
- if (targ_ns != &init_user_ns && targ_ns->creator == cred->user)
+ /* The owner of the user namespace has all caps. */
+ if (targ_ns != &init_user_ns && uid_eq(targ_ns->owner,
+ make_kuid(cred->user_ns, cred->euid)))
return 0;
/* Do we have the necessary capabilities? */
--
1.7.2.5
next prev parent reply other threads:[~2012-04-08 5:12 UTC|newest]
Thread overview: 89+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-04-08 5:10 [REVIEW][PATCH 0/43] Completing the user namespace Eric W. Biederman
2012-04-08 5:14 ` [PATCH 01/43] vfs: Don't allow a user namespace root to make device nodes "Eric W. Beiderman
2012-04-08 5:14 ` [PATCH 02/43] userns: Kill bogus declaration of function release_uids "Eric W. Beiderman
2012-04-08 5:14 ` [PATCH 03/43] userns: Replace netlink uses of cap_raised with capable "Eric W. Beiderman
2012-04-08 5:15 ` [PATCH 04/43] userns: Remove unnecessary cast to struct user_struct when copying cred->user "Eric W. Beiderman
2012-04-08 5:15 ` [PATCH 05/43] cred: Add forward declaration of init_user_ns in all cases "Eric W. Beiderman
2012-04-08 5:15 ` [PATCH 06/43] userns: Use cred->user_ns instead of cred->user->user_ns "Eric W. Beiderman
2012-04-08 5:15 ` [PATCH 07/43] cred: Refcount the user_ns pointed to by the cred "Eric W. Beiderman
2012-04-08 5:15 ` [PATCH 08/43] userns: Add an explicit reference to the parent user namespace "Eric W. Beiderman
2012-04-08 5:15 ` [PATCH 09/43] mqueue: Explicitly capture the user namespace to send the notification to "Eric W. Beiderman
2012-04-08 5:15 ` [PATCH 10/43] userns: Deprecate and rename the user_namespace reference in the user_struct "Eric W. Beiderman
2012-04-08 5:15 ` [PATCH 11/43] userns: Start out with a full set of capabilities "Eric W. Beiderman
2012-04-08 5:15 ` [PATCH 12/43] userns: Replace the hard to write inode_userns with inode_capable "Eric W. Beiderman
2012-04-08 5:15 ` [PATCH 13/43] userns: Add kuid_t and kgid_t and associated infrastructure in uidgid.h "Eric W. Beiderman
2012-04-08 5:15 ` [PATCH 14/43] userns: Add a Kconfig option to enforce strict kuid and kgid type checks "Eric W. Beiderman
2012-04-08 5:15 ` [PATCH 15/43] userns: Disassociate user_struct from the user_namespace "Eric W. Beiderman
2012-04-08 5:15 ` "Eric W. Beiderman [this message]
2012-04-18 18:48 ` [PATCH 16/43] userns: Simplify the user_namespace by making userns->creator a kuid Serge E. Hallyn
2012-04-20 22:58 ` Eric W. Biederman
2012-04-24 17:33 ` Serge E. Hallyn
2012-04-24 19:41 ` Eric W. Biederman
2012-04-24 20:23 ` Serge E. Hallyn
2012-04-26 9:09 ` Eric W. Biederman
2012-04-26 16:21 ` Serge E. Hallyn
2012-04-08 5:15 ` [PATCH 18/43] userns: Convert group_info values from gid_t to kgid_t "Eric W. Beiderman
2012-04-18 18:49 ` Serge E. Hallyn
2012-04-20 23:05 ` Eric W. Biederman
2012-04-08 5:15 ` [PATCH 19/43] userns: Store uid and gid values in struct cred with kuid_t and kgid_t types "Eric W. Beiderman
2012-04-18 18:49 ` Serge E. Hallyn
2012-04-08 5:15 ` [PATCH 20/43] userns: Replace user_ns_map_uid and user_ns_map_gid with from_kuid and from_kgid "Eric W. Beiderman
2012-04-18 18:49 ` Serge E. Hallyn
2012-04-08 5:15 ` [PATCH 21/43] userns: Convert sched_set_affinity and sched_set_scheduler's permission checks "Eric W. Beiderman
2012-04-18 18:50 ` Serge E. Hallyn
2012-04-08 5:15 ` [PATCH 22/43] userns: Convert capabilities related permsion checks "Eric W. Beiderman
2012-04-18 18:51 ` Serge E. Hallyn
2012-04-20 23:18 ` Eric W. Biederman
2012-04-08 5:15 ` [PATCH 23/43] userns: Convert setting and getting uid and gid system calls to use kuid and kgid "Eric W. Beiderman
2012-04-26 16:20 ` Serge E. Hallyn
2012-04-08 5:15 ` [PATCH 24/43] userns: Convert ptrace, kill, set_priority permission checks to work with kuids and kgids "Eric W. Beiderman
2012-04-18 18:56 ` Serge E. Hallyn
2012-04-20 23:51 ` Eric W. Biederman
2012-04-08 5:15 ` [PATCH 25/43] userns: Store uid and gid types in vfs structures with kuid_t and kgid_t types "Eric W. Beiderman
2012-04-18 18:57 ` Serge E. Hallyn
2012-04-08 5:15 ` [PATCH 26/43] userns: Convert in_group_p and in_egroup_p to use kgid_t "Eric W. Beiderman
2012-04-18 18:58 ` Serge E. Hallyn
2012-04-08 5:15 ` [PATCH 27/43] userns: Use uid_eq gid_eq helpers when comparing kuids and kgids in the vfs "Eric W. Beiderman
2012-04-18 19:02 ` Serge E. Hallyn
2012-04-21 0:05 ` Eric W. Biederman
2012-04-18 19:03 ` Serge E. Hallyn
2012-04-21 0:58 ` Eric W. Biederman
2012-04-24 17:41 ` Serge E. Hallyn
2012-04-26 0:11 ` Serge E. Hallyn
2012-04-26 5:33 ` Eric W. Biederman
2012-04-08 5:15 ` [PATCH 28/43] userns: Convert user specfied uids and gids in chown into kuids and kgid "Eric W. Beiderman
2012-04-18 19:03 ` Serge E. Hallyn
2012-04-08 5:15 ` [PATCH 29/43] userns: Convert stat to return values mapped from kuids and kgids "Eric W. Beiderman
2012-04-18 19:03 ` Serge E. Hallyn
2012-04-08 5:15 ` [PATCH 30/43] userns: Fail exec for suid and sgid binaries with ids outside our user namespace "Eric W. Beiderman
2012-04-18 19:05 ` Serge E. Hallyn
2012-04-18 19:09 ` Serge E. Hallyn
2012-04-24 2:28 ` Eric W. Biederman
2012-04-24 15:10 ` Serge Hallyn
2012-04-08 5:15 ` [PATCH 31/43] userns: Teach inode_capable to understand inodes whose uids map to other namespaces "Eric W. Beiderman
2012-04-18 19:06 ` Serge E. Hallyn
2012-04-08 5:15 ` [PATCH 32/43] userns: signal remove unnecessary map_cred_ns "Eric W. Beiderman
2012-04-18 19:07 ` Serge E. Hallyn
2012-04-08 5:15 ` [PATCH 33/43] userns: Convert binary formats to use kuid/kgid where appropriate "Eric W. Beiderman
2012-04-18 19:10 ` Serge E. Hallyn
2012-04-24 2:44 ` Eric W. Biederman
2012-04-08 5:15 ` [PATCH 34/43] userns: Convert devpts " "Eric W. Beiderman
2012-04-08 5:15 ` [PATCH 35/43] userns: Convert ext2 " "Eric W. Beiderman
2012-04-08 5:15 ` [PATCH 36/43] userns: Convert ext3 " "Eric W. Beiderman
2012-04-08 5:15 ` [PATCH 37/43] userns: Convert ext4 to user " "Eric W. Beiderman
2012-04-08 5:15 ` [PATCH 38/43] userns: Convert proc to use " "Eric W. Beiderman
2012-04-08 5:15 ` [PATCH 39/43] userns: Convert sysctl permission checks to use kuid and kgids "Eric W. Beiderman
2012-04-08 5:15 ` [PATCH 40/43] userns: Convert sysfs to use kgid/kuid where appropriate "Eric W. Beiderman
2012-04-08 5:15 ` [PATCH 41/43] userns: Convert tmpfs to use kuid and kgid " "Eric W. Beiderman
2012-04-08 5:15 ` [PATCH 42/43] userns: Convert cgroup permission checks to use uid_eq "Eric W. Beiderman
2012-04-08 5:15 ` [PATCH 43/43] userns: Convert the move_pages, and migrate_pages " "Eric W. Beiderman
2012-04-08 14:54 ` [REVIEW][PATCH 0/43] Completing the user namespace Serge Hallyn
2012-04-08 17:40 ` richard -rw- weinberger
2012-04-08 21:30 ` Eric W. Biederman
2012-04-08 22:04 ` richard -rw- weinberger
2012-04-08 22:52 ` Eric W. Biederman
[not found] ` <1333862139-31737-17-git-send-email-ebiederm@xmission.com>
2012-04-18 18:49 ` [PATCH 17/43] userns: Rework the user_namespace adding uid/gid mapping support Serge E. Hallyn
2012-05-11 23:20 ` Please include user-namespace.git in linux-next Eric W. Biederman
2012-05-13 23:35 ` Stephen Rothwell
2012-05-21 2:25 ` Tetsuo Handa
2012-05-22 17:26 ` Eric W. Biederman
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1333862139-31737-16-git-send-email-ebiederm@xmission.com \
--to=ebiederm@xmission.com \
--cc=akpm@linux-foundation.org \
--cc=containers@lists.linux-foundation.org \
--cc=gorcunov@openvz.org \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-security-module@vger.kernel.org \
--cc=torvalds@linux-foundation.org \
--cc=viro@ZenIV.linux.org.uk \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®