From: Manfred Spraul <manfred@colorfullife.com>
To: Ethan Zhao <ethan.kernel@gmail.com>
Cc: Ethan Zhao <ethan.zhao@oracle.com>,
LKML <linux-kernel@vger.kernel.org>,
linux-security-module@vger.kernel.org, james.l.morris@oracle.com,
Stephen Smalley <sds@tycho.nsa.gov>,
selinux@tycho.nsa.gov
Subject: Re: [PATCH] Selinux/hooks.c: Fix a NULL pointer dereference caused by semop()
Date: Thu, 22 Jan 2015 19:15:55 +0100 [thread overview]
Message-ID: <54C13E5B.3020208@colorfullife.com> (raw)
In-Reply-To: <CABawtvO3PdKhmeH_H1v_tQxmgOBQrnLGNayTE1=EbcxJbtwAzQ@mail.gmail.com>
[-- Attachment #1: Type: text/plain, Size: 2726 bytes --]
On 01/22/2015 03:44 AM, Ethan Zhao wrote:
> On Wed, Jan 21, 2015 at 1:30 PM, Manfred Spraul
> <manfred@colorfullife.com> wrote:
>> On 01/21/2015 04:53 AM, Ethan Zhao wrote:
>>> On Tue, Jan 20, 2015 at 10:10 PM, Stephen Smalley <sds@tycho.nsa.gov>
>>> wrote:
>>>> On 01/20/2015 04:18 AM, Ethan Zhao wrote:
>>>>> sys_semget()
>>>>> ->newary()
>>>>> ->security_sem_alloc()
>>>>> ->sem_alloc_security()
>>>>> selinux_sem_alloc_security()
>>>>> ->ipc_alloc_security() {
>>>>> ->rc = avc_has_perm()
>>>>> if (rc) {
>>>>>
>>>>> ipc_free_security(&sma->sem_perm);
>>>>> return rc;
>>>> We free the security structure here to avoid a memory leak on a
>>>> failed/denied semaphore set creation. In this situation, we return an
>>>> error to the caller (ultimately to newary), it does an
>>>> ipc_rcu_putref(sma, ipc_rcu_free), and it returns an error to the
>>>> caller. Thus, it never calls ipc_addid() and the semaphore set is not
>>>> created. So how then can you call semtimedop() on it?
>>> Seems it wouldn't happen after commit
>>> e8577d1f0329d4842e8302e289fb2c22156abef4 ?
>> That was my first guess when I read the bug report - but it can't be the
>> fix, because security_sem_alloc() is before the ipc_addid(), with or without
>> the patch.
>>
>> thread A:
>> thread B:
>>
>> semtimedop()
>> -> sem_obtain_object_check()
>> semctl(IPC_RMID)
>> -> freeary()
>> -> ipc_rcu_putref()
>> -> call_rcu()
>> -> somehow a grace period
>> -> sem_rcu_free()
>> -> security_sem_free()
>>
>> Perhaps: modify ipc_free_security() to hexdump perm and a few more bytes if
>> the pointer is NULL?
> I tried to ask for vmcore and do more analysis, basically, the race condition
> still exists and open a hole to be DoS.
Is the issue reproducable?
If yes, can you try something like the attached patch?
> --- a/security/selinux/hooks.c
> +++ b/security/selinux/hooks.c
> @@ -5129,6 +5129,8 @@ static int ipc_has_perm(struct kern_ipc_perm
> *ipc_perms,
> u32 sid = current_sid();
>
> isec = ipc_perms->security;
> + if (!isec)
> + return -EACCES;
>
> ad.type = LSM_AUDIT_DATA_IPC;
> ad.u.ipc_id = ipc_perms->key;
ipc_has_perm() runs without any spinlocks/semaphores, only rcu_read_lock().
Testing for ipc_perms->security!=NULL does solve the issue that
ipc_perm->key could be an access to kfree'd memory: Nothing prevents
that the kfree could happen just after the test.
I.e.: The patch can't be the right solution.
--
Manfred
[-- Attachment #2: patch-sem_ipc_has_perm --]
[-- Type: text/plain, Size: 2351 bytes --]
diff --git a/ipc/sem.c b/ipc/sem.c
index 6115146..80371dc 100644
--- a/ipc/sem.c
+++ b/ipc/sem.c
@@ -248,6 +248,7 @@ static void sem_rcu_free(struct rcu_head *head)
struct ipc_rcu *p = container_of(head, struct ipc_rcu, rcu);
struct sem_array *sma = ipc_rcu_to_struct(p);
+pr_info("sem_rcu_free: sma %p\n",sma);
security_sem_free(sma);
ipc_rcu_free(head);
}
@@ -529,6 +530,7 @@ static int newary(struct ipc_namespace *ns, struct ipc_params *params)
sma->sem_nsems = nsems;
sma->sem_ctime = get_seconds();
+pr_info("newary: sma %p becomes visible\n",sma);
id = ipc_addid(&sem_ids(ns), &sma->sem_perm, ns->sc_semmni);
if (id < 0) {
ipc_rcu_putref(sma, sem_rcu_free);
@@ -1118,6 +1120,7 @@ static void freeary(struct ipc_namespace *ns, struct kern_ipc_perm *ipcp)
/* Remove the semaphore set from the IDR */
sem_rmid(ns, sma);
+pr_info("freeary: sma %p unlinked\n",sma);
sem_unlock(sma, -1);
rcu_read_unlock();
@@ -1860,6 +1863,9 @@ SYSCALL_DEFINE4(semtimedop, int, semid, struct sembuf __user *, tsops,
if (ipcperms(ns, &sma->sem_perm, alter ? S_IWUGO : S_IRUGO))
goto out_rcu_wakeup;
+ if (sma->sem_perm.security == NULL) {
+ pr_info("sma %p: sem_perm.security == NULL\n", sma);
+ }
error = security_sem_semop(sma, sops, nsops, alter);
if (error)
goto out_rcu_wakeup;
diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c
index 6da7532..1499787 100644
--- a/security/selinux/hooks.c
+++ b/security/selinux/hooks.c
@@ -5088,6 +5088,7 @@ static int ipc_alloc_security(struct task_struct *task,
isec->sclass = sclass;
isec->sid = sid;
perm->security = isec;
+pr_info("ipc_alloc_security for perm %p.\n", perm);
return 0;
}
@@ -5096,6 +5097,7 @@ static void ipc_free_security(struct kern_ipc_perm *perm)
{
struct ipc_security_struct *isec = perm->security;
perm->security = NULL;
+pr_info("ipc_free_security for perm %p.\n", perm);
kfree(isec);
}
@@ -5129,6 +5131,12 @@ static int ipc_has_perm(struct kern_ipc_perm *ipc_perms,
u32 sid = current_sid();
isec = ipc_perms->security;
+ if (isec == NULL) {
+ struct sem_array *sma = container_of(ipc_perms, struct sem_array, sem_perm);
+
+ pr_info("sma %p, sem_base %p, deleted %d with NULL isec\n",
+ sma, sma->sem_base, sma->sem_perm.deleted);
+ }
ad.type = LSM_AUDIT_DATA_IPC;
ad.u.ipc_id = ipc_perms->key;
WARNING: multiple messages have this Message-ID (diff)
From: Manfred Spraul <manfred@colorfullife.com>
To: Ethan Zhao <ethan.kernel@gmail.com>
Cc: Stephen Smalley <sds@tycho.nsa.gov>,
Ethan Zhao <ethan.zhao@oracle.com>,
james.l.morris@oracle.com, serge@hallyn.com,
Eric Paris <eparis@parisplace.org>,
Paul Moore <paul@paul-moore.com>,
selinux@tycho.nsa.gov, linux-security-module@vger.kernel.org,
LKML <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH] Selinux/hooks.c: Fix a NULL pointer dereference caused by semop()
Date: Thu, 22 Jan 2015 19:15:55 +0100 [thread overview]
Message-ID: <54C13E5B.3020208@colorfullife.com> (raw)
In-Reply-To: <CABawtvO3PdKhmeH_H1v_tQxmgOBQrnLGNayTE1=EbcxJbtwAzQ@mail.gmail.com>
[-- Attachment #1: Type: text/plain, Size: 2726 bytes --]
On 01/22/2015 03:44 AM, Ethan Zhao wrote:
> On Wed, Jan 21, 2015 at 1:30 PM, Manfred Spraul
> <manfred@colorfullife.com> wrote:
>> On 01/21/2015 04:53 AM, Ethan Zhao wrote:
>>> On Tue, Jan 20, 2015 at 10:10 PM, Stephen Smalley <sds@tycho.nsa.gov>
>>> wrote:
>>>> On 01/20/2015 04:18 AM, Ethan Zhao wrote:
>>>>> sys_semget()
>>>>> ->newary()
>>>>> ->security_sem_alloc()
>>>>> ->sem_alloc_security()
>>>>> selinux_sem_alloc_security()
>>>>> ->ipc_alloc_security() {
>>>>> ->rc = avc_has_perm()
>>>>> if (rc) {
>>>>>
>>>>> ipc_free_security(&sma->sem_perm);
>>>>> return rc;
>>>> We free the security structure here to avoid a memory leak on a
>>>> failed/denied semaphore set creation. In this situation, we return an
>>>> error to the caller (ultimately to newary), it does an
>>>> ipc_rcu_putref(sma, ipc_rcu_free), and it returns an error to the
>>>> caller. Thus, it never calls ipc_addid() and the semaphore set is not
>>>> created. So how then can you call semtimedop() on it?
>>> Seems it wouldn't happen after commit
>>> e8577d1f0329d4842e8302e289fb2c22156abef4 ?
>> That was my first guess when I read the bug report - but it can't be the
>> fix, because security_sem_alloc() is before the ipc_addid(), with or without
>> the patch.
>>
>> thread A:
>> thread B:
>>
>> semtimedop()
>> -> sem_obtain_object_check()
>> semctl(IPC_RMID)
>> -> freeary()
>> -> ipc_rcu_putref()
>> -> call_rcu()
>> -> somehow a grace period
>> -> sem_rcu_free()
>> -> security_sem_free()
>>
>> Perhaps: modify ipc_free_security() to hexdump perm and a few more bytes if
>> the pointer is NULL?
> I tried to ask for vmcore and do more analysis, basically, the race condition
> still exists and open a hole to be DoS.
Is the issue reproducable?
If yes, can you try something like the attached patch?
> --- a/security/selinux/hooks.c
> +++ b/security/selinux/hooks.c
> @@ -5129,6 +5129,8 @@ static int ipc_has_perm(struct kern_ipc_perm
> *ipc_perms,
> u32 sid = current_sid();
>
> isec = ipc_perms->security;
> + if (!isec)
> + return -EACCES;
>
> ad.type = LSM_AUDIT_DATA_IPC;
> ad.u.ipc_id = ipc_perms->key;
ipc_has_perm() runs without any spinlocks/semaphores, only rcu_read_lock().
Testing for ipc_perms->security!=NULL does solve the issue that
ipc_perm->key could be an access to kfree'd memory: Nothing prevents
that the kfree could happen just after the test.
I.e.: The patch can't be the right solution.
--
Manfred
[-- Attachment #2: patch-sem_ipc_has_perm --]
[-- Type: text/plain, Size: 2351 bytes --]
diff --git a/ipc/sem.c b/ipc/sem.c
index 6115146..80371dc 100644
--- a/ipc/sem.c
+++ b/ipc/sem.c
@@ -248,6 +248,7 @@ static void sem_rcu_free(struct rcu_head *head)
struct ipc_rcu *p = container_of(head, struct ipc_rcu, rcu);
struct sem_array *sma = ipc_rcu_to_struct(p);
+pr_info("sem_rcu_free: sma %p\n",sma);
security_sem_free(sma);
ipc_rcu_free(head);
}
@@ -529,6 +530,7 @@ static int newary(struct ipc_namespace *ns, struct ipc_params *params)
sma->sem_nsems = nsems;
sma->sem_ctime = get_seconds();
+pr_info("newary: sma %p becomes visible\n",sma);
id = ipc_addid(&sem_ids(ns), &sma->sem_perm, ns->sc_semmni);
if (id < 0) {
ipc_rcu_putref(sma, sem_rcu_free);
@@ -1118,6 +1120,7 @@ static void freeary(struct ipc_namespace *ns, struct kern_ipc_perm *ipcp)
/* Remove the semaphore set from the IDR */
sem_rmid(ns, sma);
+pr_info("freeary: sma %p unlinked\n",sma);
sem_unlock(sma, -1);
rcu_read_unlock();
@@ -1860,6 +1863,9 @@ SYSCALL_DEFINE4(semtimedop, int, semid, struct sembuf __user *, tsops,
if (ipcperms(ns, &sma->sem_perm, alter ? S_IWUGO : S_IRUGO))
goto out_rcu_wakeup;
+ if (sma->sem_perm.security == NULL) {
+ pr_info("sma %p: sem_perm.security == NULL\n", sma);
+ }
error = security_sem_semop(sma, sops, nsops, alter);
if (error)
goto out_rcu_wakeup;
diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c
index 6da7532..1499787 100644
--- a/security/selinux/hooks.c
+++ b/security/selinux/hooks.c
@@ -5088,6 +5088,7 @@ static int ipc_alloc_security(struct task_struct *task,
isec->sclass = sclass;
isec->sid = sid;
perm->security = isec;
+pr_info("ipc_alloc_security for perm %p.\n", perm);
return 0;
}
@@ -5096,6 +5097,7 @@ static void ipc_free_security(struct kern_ipc_perm *perm)
{
struct ipc_security_struct *isec = perm->security;
perm->security = NULL;
+pr_info("ipc_free_security for perm %p.\n", perm);
kfree(isec);
}
@@ -5129,6 +5131,12 @@ static int ipc_has_perm(struct kern_ipc_perm *ipc_perms,
u32 sid = current_sid();
isec = ipc_perms->security;
+ if (isec == NULL) {
+ struct sem_array *sma = container_of(ipc_perms, struct sem_array, sem_perm);
+
+ pr_info("sma %p, sem_base %p, deleted %d with NULL isec\n",
+ sma, sma->sem_base, sma->sem_perm.deleted);
+ }
ad.type = LSM_AUDIT_DATA_IPC;
ad.u.ipc_id = ipc_perms->key;
next prev parent reply other threads:[~2015-01-22 18:16 UTC|newest]
Thread overview: 35+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-01-20 9:18 [PATCH] Selinux/hooks.c: Fix a NULL pointer dereference caused by semop() Ethan Zhao
2015-01-20 9:18 ` Ethan Zhao
2015-01-20 14:10 ` Stephen Smalley
2015-01-20 14:10 ` Stephen Smalley
2015-01-20 18:49 ` Manfred Spraul
2015-01-20 18:49 ` Manfred Spraul
2015-01-20 21:01 ` Stephen Smalley
2015-01-20 21:06 ` Eric Paris
2015-01-20 21:06 ` Eric Paris
2015-01-20 21:09 ` Stephen Smalley
2015-01-20 21:09 ` Stephen Smalley
2015-01-21 1:30 ` ethan zhao
2015-01-21 1:30 ` ethan zhao
2015-01-21 3:53 ` Ethan Zhao
2015-01-21 3:53 ` Ethan Zhao
2015-01-21 5:30 ` Manfred Spraul
2015-01-21 5:30 ` Manfred Spraul
2015-01-22 2:44 ` Ethan Zhao
2015-01-22 2:44 ` Ethan Zhao
2015-01-22 18:15 ` Manfred Spraul [this message]
2015-01-22 18:15 ` Manfred Spraul
2015-01-23 2:00 ` ethan zhao
2015-01-23 2:00 ` ethan zhao
2015-01-22 19:05 ` Stephen Smalley
2015-01-22 19:05 ` Stephen Smalley
2015-01-22 20:48 ` Davidlohr Bueso
2015-01-22 20:48 ` Davidlohr Bueso
2015-01-23 2:38 ` ethan zhao
2015-01-23 2:38 ` ethan zhao
2015-01-23 2:19 ` ethan zhao
2015-01-23 2:19 ` ethan zhao
2015-01-23 3:30 ` Davidlohr Bueso
2015-01-23 3:30 ` Davidlohr Bueso
2015-01-23 15:30 ` Ethan Zhao
2015-01-23 15:30 ` Ethan Zhao
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=54C13E5B.3020208@colorfullife.com \
--to=manfred@colorfullife.com \
--cc=ethan.kernel@gmail.com \
--cc=ethan.zhao@oracle.com \
--cc=james.l.morris@oracle.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-security-module@vger.kernel.org \
--cc=sds@tycho.nsa.gov \
--cc=selinux@tycho.nsa.gov \
/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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.