The patch titled
userns: user namespaces: convert several capable() calls
has been added to the -mm tree. Its filename is
userns-user-namespaces-convert-several-capable-calls.patch
Before you just go and hit "reply", please:
a) Consider who else should be cc'ed
b) Prefer to cc a suitable mailing list as well
c) Ideally: find the original patch on the mailing list and do a
reply-to-all to that, adding suitable additional cc's
*** Remember to use Documentation/SubmitChecklist when testing your code ***
See http://userweb.kernel.org/~akpm/stuff/added-to-mm.txt to find
out what to do about this
The current -mm tree may be found at http://userweb.kernel.org/~akpm/mmotm/
------------------------------------------------------
Subject: userns: user namespaces: convert several capable() calls
From: "Serge E. Hallyn" <serge@xxxxxxxxxx>
CAP_IPC_OWNER and CAP_IPC_LOCK can be checked against current_user_ns(),
because the resource comes from current's own ipc namespace.
setuid/setgid are to uids in own namespace, so again checks can be against
current_user_ns().
Changelog:
Jan 11: Use task_ns_capable() in place of sched_capable().
Jan 11: Use nsown_capable() as suggested by Bastian Blank.
Jan 11: Clarify (hopefully) some logic in futex and sched.c
Feb 15: use ns_capable for ipc, not nsown_capable
Feb 23: let copy_ipcs handle setting ipc_ns->user_ns
Feb 23: pass ns down rather than taking it from current
Signed-off-by: Serge E. Hallyn <serge.hallyn@xxxxxxxxxxxxx>
Acked-by: "Eric W. Biederman" <ebiederm@xxxxxxxxxxxx>
Acked-by: Daniel Lezcano <daniel.lezcano@xxxxxxx>
Acked-by: David Howells <dhowells@xxxxxxxxxx>
Cc: James Morris <jmorris@xxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
include/linux/ipc_namespace.h | 7 ++++---
ipc/msg.c | 8 ++++----
ipc/namespace.c | 13 ++++++++-----
ipc/sem.c | 9 +++++----
ipc/shm.c | 9 +++++----
ipc/util.c | 22 +++++++++++++---------
ipc/util.h | 5 +++--
kernel/futex.c | 11 ++++++++++-
kernel/futex_compat.c | 11 ++++++++++-
kernel/groups.c | 2 +-
kernel/nsproxy.c | 7 +------
kernel/sched.c | 9 ++++++---
kernel/uid16.c | 2 +-
13 files changed, 71 insertions(+), 44 deletions(-)
diff -puN include/linux/ipc_namespace.h~userns-user-namespaces-convert-several-capable-calls include/linux/ipc_namespace.h
--- a/include/linux/ipc_namespace.h~userns-user-namespaces-convert-several-capable-calls
+++ a/include/linux/ipc_namespace.h
@@ -5,6 +5,7 @@
#include <linux/idr.h>
#include <linux/rwsem.h>
#include <linux/notifier.h>
+#include <linux/nsproxy.h>
/*
* ipc namespace events
@@ -94,7 +95,7 @@ static inline int mq_init_ns(struct ipc_
#if defined(CONFIG_IPC_NS)
extern struct ipc_namespace *copy_ipcs(unsigned long flags,
- struct ipc_namespace *ns);
+ struct task_struct *tsk);
static inline struct ipc_namespace *get_ipc_ns(struct ipc_namespace *ns)
{
if (ns)
@@ -105,12 +106,12 @@ static inline struct ipc_namespace *get_
extern void put_ipc_ns(struct ipc_namespace *ns);
#else
static inline struct ipc_namespace *copy_ipcs(unsigned long flags,
- struct ipc_namespace *ns)
+ struct task_struct *tsk)
{
if (flags & CLONE_NEWIPC)
return ERR_PTR(-EINVAL);
- return ns;
+ return tsk->nsproxy->ipc_ns;
}
static inline struct ipc_namespace *get_ipc_ns(struct ipc_namespace *ns)
diff -puN ipc/msg.c~userns-user-namespaces-convert-several-capable-calls ipc/msg.c
--- a/ipc/msg.c~userns-user-namespaces-convert-several-capable-calls
+++ a/ipc/msg.c
@@ -421,7 +421,7 @@ static int msgctl_down(struct ipc_namesp
return -EFAULT;
}
- ipcp = ipcctl_pre_down(&msg_ids(ns), msqid, cmd,
+ ipcp = ipcctl_pre_down(ns, &msg_ids(ns), msqid, cmd,
&msqid64.msg_perm, msqid64.msg_qbytes);
if (IS_ERR(ipcp))
return PTR_ERR(ipcp);
@@ -539,7 +539,7 @@ SYSCALL_DEFINE3(msgctl, int, msqid, int,
success_return = 0;
}
err = -EACCES;
- if (ipcperms(&msq->q_perm, S_IRUGO))
+ if (ipcperms(ns, &msq->q_perm, S_IRUGO))
goto out_unlock;
err = security_msg_queue_msgctl(msq, cmd);
@@ -664,7 +664,7 @@ long do_msgsnd(int msqid, long mtype, vo
struct msg_sender s;
err = -EACCES;
- if (ipcperms(&msq->q_perm, S_IWUGO))
+ if (ipcperms(ns, &msq->q_perm, S_IWUGO))
goto out_unlock_free;
err = security_msg_queue_msgsnd(msq, msg, msgflg);
@@ -774,7 +774,7 @@ long do_msgrcv(int msqid, long *pmtype,
struct list_head *tmp;
msg = ERR_PTR(-EACCES);
- if (ipcperms(&msq->q_perm, S_IRUGO))
+ if (ipcperms(ns, &msq->q_perm, S_IRUGO))
goto out_unlock;
msg = ERR_PTR(-EAGAIN);
diff -puN ipc/namespace.c~userns-user-namespaces-convert-several-capable-calls ipc/namespace.c
--- a/ipc/namespace.c~userns-user-namespaces-convert-several-capable-calls
+++ a/ipc/namespace.c
@@ -15,7 +15,8 @@
#include "util.h"
-static struct ipc_namespace *create_ipc_ns(struct ipc_namespace *old_ns)
+static struct ipc_namespace *create_ipc_ns(struct task_struct *tsk,
+ struct ipc_namespace *old_ns)
{
struct ipc_namespace *ns;
int err;
@@ -44,17 +45,19 @@ static struct ipc_namespace *create_ipc_
ipcns_notify(IPCNS_CREATED);
register_ipcns_notifier(ns);
- ns->user_ns = old_ns->user_ns;
- get_user_ns(ns->user_ns);
+ ns->user_ns = get_user_ns(task_cred_xxx(tsk, user)->user_ns);
return ns;
}
-struct ipc_namespace *copy_ipcs(unsigned long flags, struct ipc_namespace *ns)
+struct ipc_namespace *copy_ipcs(unsigned long flags,
+ struct task_struct *tsk)
{
+ struct ipc_namespace *ns = tsk->nsproxy->ipc_ns;
+
if (!(flags & CLONE_NEWIPC))
return get_ipc_ns(ns);
- return create_ipc_ns(ns);
+ return create_ipc_ns(tsk, ns);
}
/*
diff -puN ipc/sem.c~userns-user-namespaces-convert-several-capable-calls ipc/sem.c
--- a/ipc/sem.c~userns-user-namespaces-convert-several-capable-calls
+++ a/ipc/sem.c
@@ -817,7 +817,7 @@ static int semctl_nolock(struct ipc_name
}
err = -EACCES;
- if (ipcperms (&sma->sem_perm, S_IRUGO))
+ if (ipcperms (ns, &sma->sem_perm, S_IRUGO))
goto out_unlock;
err = security_sem_semctl(sma, cmd);
@@ -862,7 +862,7 @@ static int semctl_main(struct ipc_namesp
nsems = sma->sem_nsems;
err = -EACCES;
- if (ipcperms (&sma->sem_perm, (cmd==SETVAL||cmd==SETALL)?S_IWUGO:S_IRUGO))
+ if (ipcperms (ns, &sma->sem_perm, (cmd==SETVAL||cmd==SETALL)?S_IWUGO:S_IRUGO))
goto out_unlock;
err = security_sem_semctl(sma, cmd);
@@ -1047,7 +1047,8 @@ static int semctl_down(struct ipc_namesp
return -EFAULT;
}
- ipcp = ipcctl_pre_down(&sem_ids(ns), semid, cmd, &semid64.sem_perm, 0);
+ ipcp = ipcctl_pre_down(ns, &sem_ids(ns), semid, cmd,
+ &semid64.sem_perm, 0);
if (IS_ERR(ipcp))
return PTR_ERR(ipcp);
@@ -1386,7 +1387,7 @@ SYSCALL_DEFINE4(semtimedop, int, semid,
goto out_unlock_free;
error = -EACCES;
- if (ipcperms(&sma->sem_perm, alter ? S_IWUGO : S_IRUGO))
+ if (ipcperms(ns, &sma->sem_perm, alter ? S_IWUGO : S_IRUGO))
goto out_unlock_free;
error = security_sem_semop(sma, sops, nsops, alter);
diff -puN ipc/shm.c~userns-user-namespaces-convert-several-capable-calls ipc/shm.c
--- a/ipc/shm.c~userns-user-namespaces-convert-several-capable-calls
+++ a/ipc/shm.c
@@ -623,7 +623,8 @@ static int shmctl_down(struct ipc_namesp
return -EFAULT;
}
- ipcp = ipcctl_pre_down(&shm_ids(ns), shmid, cmd, &shmid64.shm_perm, 0);
+ ipcp = ipcctl_pre_down(ns, &shm_ids(ns), shmid, cmd,
+ &shmid64.shm_perm, 0);
if (IS_ERR(ipcp))
return PTR_ERR(ipcp);
@@ -737,7 +738,7 @@ SYSCALL_DEFINE3(shmctl, int, shmid, int,
result = 0;
}
err = -EACCES;
- if (ipcperms (&shp->shm_perm, S_IRUGO))
+ if (ipcperms (ns, &shp->shm_perm, S_IRUGO))
goto out_unlock;
err = security_shm_shmctl(shp, cmd);
if (err)
@@ -773,7 +774,7 @@ SYSCALL_DEFINE3(shmctl, int, shmid, int,
audit_ipc_obj(&(shp->shm_perm));
- if (!capable(CAP_IPC_LOCK)) {
+ if (!ns_capable(ns->user_ns, CAP_IPC_LOCK)) {
uid_t euid = current_euid();
err = -EPERM;
if (euid != shp->shm_perm.uid &&
@@ -888,7 +889,7 @@ long do_shmat(int shmid, char __user *sh
}
err = -EACCES;
- if (ipcperms(&shp->shm_perm, acc_mode))
+ if (ipcperms(ns, &shp->shm_perm, acc_mode))
goto out_unlock;
err = security_shm_shmat(shp, shmaddr, shmflg);
diff -puN ipc/util.c~userns-user-namespaces-convert-several-capable-calls ipc/util.c
--- a/ipc/util.c~userns-user-namespaces-convert-several-capable-calls
+++ a/ipc/util.c
@@ -329,12 +329,14 @@ retry:
*
* It is called with ipc_ids.rw_mutex and ipcp->lock held.
*/
-static int ipc_check_perms(struct kern_ipc_perm *ipcp, struct ipc_ops *ops,
- struct ipc_params *params)
+static int ipc_check_perms(struct ipc_namespace *ns,
+ struct kern_ipc_perm *ipcp,
+ struct ipc_ops *ops,
+ struct ipc_params *params)
{
int err;
- if (ipcperms(ipcp, params->flg))
+ if (ipcperms(ns, ipcp, params->flg))
err = -EACCES;
else {
err = ops->associate(ipcp, params->flg);
@@ -396,7 +398,7 @@ retry:
* ipc_check_perms returns the IPC id on
* success
*/
- err = ipc_check_perms(ipcp, ops, params);
+ err = ipc_check_perms(ns, ipcp, ops, params);
}
ipc_unlock(ipcp);
}
@@ -612,7 +614,7 @@ void ipc_rcu_putref(void *ptr)
* to ipc resources. return 0 if allowed
*/
-int ipcperms (struct kern_ipc_perm *ipcp, short flag)
+int ipcperms (struct ipc_namespace *ns, struct kern_ipc_perm *ipcp, short flag)
{ /* flag will most probably be 0 or S_...UGO from <linux/stat.h> */
uid_t euid = current_euid();
int requested_mode, granted_mode;
@@ -627,7 +629,7 @@ int ipcperms (struct kern_ipc_perm *ipcp
granted_mode >>= 3;
/* is there some bit set in requested_mode but not in granted_mode? */
if ((requested_mode & ~granted_mode & 0007) &&
- !capable(CAP_IPC_OWNER))
+ !ns_capable(ns->user_ns, CAP_IPC_OWNER))
return -1;
return security_ipc_permission(ipcp, flag);
@@ -765,6 +767,7 @@ void ipc_update_perm(struct ipc64_perm *
/**
* ipcctl_pre_down - retrieve an ipc and check permissions for some IPC_XXX cmd
+ * @ids: the ipc namespace
* @ids: the table of ids where to look for the ipc
* @id: the id of the ipc to retrieve
* @cmd: the cmd to check
@@ -779,7 +782,8 @@ void ipc_update_perm(struct ipc64_perm *
* - returns the ipc with both ipc and rw_mutex locks held in case of success
* or an err-code without any lock held otherwise.
*/
-struct kern_ipc_perm *ipcctl_pre_down(struct ipc_ids *ids, int id, int cmd,
+struct kern_ipc_perm *ipcctl_pre_down(struct ipc_namespace *ns,
+ struct ipc_ids *ids, int id, int cmd,
struct ipc64_perm *perm, int extra_perm)
{
struct kern_ipc_perm *ipcp;
@@ -799,8 +803,8 @@ struct kern_ipc_perm *ipcctl_pre_down(st
perm->gid, perm->mode);
euid = current_euid();
- if (euid == ipcp->cuid ||
- euid == ipcp->uid || capable(CAP_SYS_ADMIN))
+ if (euid == ipcp->cuid || euid == ipcp->uid ||
+ ns_capable(ns->user_ns, CAP_SYS_ADMIN))
return ipcp;
err = -EPERM;
diff -puN ipc/util.h~userns-user-namespaces-convert-several-capable-calls ipc/util.h
--- a/ipc/util.h~userns-user-namespaces-convert-several-capable-calls
+++ a/ipc/util.h
@@ -103,7 +103,7 @@ int ipc_get_maxid(struct ipc_ids *);
void ipc_rmid(struct ipc_ids *, struct kern_ipc_perm *);
/* must be called with ipcp locked */
-int ipcperms(struct kern_ipc_perm *ipcp, short flg);
+int ipcperms(struct ipc_namespace *ns, struct kern_ipc_perm *ipcp, short flg);
/* for rare, potentially huge allocations.
* both function can sleep
@@ -126,7 +126,8 @@ struct kern_ipc_perm *ipc_lock(struct ip
void kernel_to_ipc64_perm(struct kern_ipc_perm *in, struct ipc64_perm *out);
void ipc64_perm_to_ipc_perm(struct ipc64_perm *in, struct ipc_perm *out);
void ipc_update_perm(struct ipc64_perm *in, struct kern_ipc_perm *out);
-struct kern_ipc_perm *ipcctl_pre_down(struct ipc_ids *ids, int id, int cmd,
+struct kern_ipc_perm *ipcctl_pre_down(struct ipc_namespace *ns,
+ struct ipc_ids *ids, int id, int cmd,
struct ipc64_perm *perm, int extra_perm);
#ifndef __ARCH_WANT_IPC_PARSE_VERSION
diff -puN kernel/futex.c~userns-user-namespaces-convert-several-capable-calls kernel/futex.c
--- a/kernel/futex.c~userns-user-namespaces-convert-several-capable-calls
+++ a/kernel/futex.c
@@ -2423,10 +2423,19 @@ SYSCALL_DEFINE3(get_robust_list, int, pi
goto err_unlock;
ret = -EPERM;
pcred = __task_cred(p);
+ /* If victim is in different user_ns, then uids are not
+ comparable, so we must have CAP_SYS_PTRACE */
+ if (cred->user->user_ns != pcred->user->user_ns) {
+ if (!ns_capable(pcred->user->user_ns, CAP_SYS_PTRACE))
+ goto err_unlock;
+ goto ok;
+ }
+ /* If victim is in same user_ns, then uids are comparable */
if (cred->euid != pcred->euid &&
cred->euid != pcred->uid &&
- !capable(CAP_SYS_PTRACE))
+ !ns_capable(pcred->user->user_ns, CAP_SYS_PTRACE))
goto err_unlock;
+ok:
head = p->robust_list;
rcu_read_unlock();
}
diff -puN kernel/futex_compat.c~userns-user-namespaces-convert-several-capable-calls kernel/futex_compat.c
--- a/kernel/futex_compat.c~userns-user-namespaces-convert-several-capable-calls
+++ a/kernel/futex_compat.c
@@ -153,10 +153,19 @@ compat_sys_get_robust_list(int pid, comp
goto err_unlock;
ret = -EPERM;
pcred = __task_cred(p);
+ /* If victim is in different user_ns, then uids are not
+ comparable, so we must have CAP_SYS_PTRACE */
+ if (cred->user->user_ns != pcred->user->user_ns) {
+ if (!ns_capable(pcred->user->user_ns, CAP_SYS_PTRACE))
+ goto err_unlock;
+ goto ok;
+ }
+ /* If victim is in same user_ns, then uids are comparable */
if (cred->euid != pcred->euid &&
cred->euid != pcred->uid &&
- !capable(CAP_SYS_PTRACE))
+ !ns_capable(pcred->user->user_ns, CAP_SYS_PTRACE))
goto err_unlock;
+ok:
head = p->compat_robust_list;
rcu_read_unlock();
}
diff -puN kernel/groups.c~userns-user-namespaces-convert-several-capable-calls kernel/groups.c
--- a/kernel/groups.c~userns-user-namespaces-convert-several-capable-calls
+++ a/kernel/groups.c
@@ -233,7 +233,7 @@ SYSCALL_DEFINE2(setgroups, int, gidsetsi
struct group_info *group_info;
int retval;
- if (!capable(CAP_SETGID))
+ if (!nsown_capable(CAP_SETGID))
return -EPERM;
if ((unsigned)gidsetsize > NGROUPS_MAX)
return -EINVAL;
diff -puN kernel/nsproxy.c~userns-user-namespaces-convert-several-capable-calls kernel/nsproxy.c
--- a/kernel/nsproxy.c~userns-user-namespaces-convert-several-capable-calls
+++ a/kernel/nsproxy.c
@@ -75,16 +75,11 @@ static struct nsproxy *create_new_namesp
goto out_uts;
}
- new_nsp->ipc_ns = copy_ipcs(flags, tsk->nsproxy->ipc_ns);
+ new_nsp->ipc_ns = copy_ipcs(flags, tsk);
if (IS_ERR(new_nsp->ipc_ns)) {
err = PTR_ERR(new_nsp->ipc_ns);
goto out_ipc;
}
- if (new_nsp->ipc_ns != tsk->nsproxy->ipc_ns) {
- put_user_ns(new_nsp->ipc_ns->user_ns);
- new_nsp->ipc_ns->user_ns = task_cred_xxx(tsk, user)->user_ns;
- get_user_ns(new_nsp->ipc_ns->user_ns);
- }
new_nsp->pid_ns = copy_pid_ns(flags, task_active_pid_ns(tsk));
if (IS_ERR(new_nsp->pid_ns)) {
diff -puN kernel/sched.c~userns-user-namespaces-convert-several-capable-calls kernel/sched.c
--- a/kernel/sched.c~userns-user-namespaces-convert-several-capable-calls
+++ a/kernel/sched.c
@@ -4899,8 +4899,11 @@ static bool check_same_owner(struct task
rcu_read_lock();
pcred = __task_cred(p);
- match = (cred->euid == pcred->euid ||
- cred->euid == pcred->uid);
+ if (cred->user->user_ns == pcred->user->user_ns)
+ match = (cred->euid == pcred->euid ||
+ cred->euid == pcred->uid);
+ else
+ match = false;
rcu_read_unlock();
return match;
}
@@ -5225,7 +5228,7 @@ long sched_setaffinity(pid_t pid, const
goto out_free_cpus_allowed;
}
retval = -EPERM;
- if (!check_same_owner(p) && !capable(CAP_SYS_NICE))
+ if (!check_same_owner(p) && !task_ns_capable(p, CAP_SYS_NICE))
goto out_unlock;
retval = security_task_setscheduler(p);
diff -puN kernel/uid16.c~userns-user-namespaces-convert-several-capable-calls kernel/uid16.c
--- a/kernel/uid16.c~userns-user-namespaces-convert-several-capable-calls
+++ a/kernel/uid16.c
@@ -189,7 +189,7 @@ SYSCALL_DEFINE2(setgroups16, int, gidset
struct group_info *group_info;
int retval;
- if (!capable(CAP_SETGID))
+ if (!nsown_capable(CAP_SETGID))
return -EPERM;
if ((unsigned)gidsetsize > NGROUPS_MAX)
return -EINVAL;
_
Patches currently in -mm which might be from serge@xxxxxxxxxx are
lib-hexdumpc-make-hex2bin-return-the-updated-src-address.patch
fs-binfmt_miscc-use-kernels-hex_to_bin-method.patch
fs-binfmt_miscc-use-kernels-hex_to_bin-method-fix.patch
fs-binfmt_miscc-use-kernels-hex_to_bin-method-fix-fix.patch
pid-remove-the-child_reaper-special-case-in-init-mainc.patch
pidns-call-pid_ns_prepare_proc-from-create_pid_namespace.patch
procfs-kill-the-global-proc_mnt-variable.patch
userns-add-a-user_namespace-as-creator-owner-of-uts_namespace.patch
userns-security-make-capabilities-relative-to-the-user-namespace.patch
userns-allow-sethostname-in-a-container.patch
userns-allow-killing-tasks-in-your-own-or-child-userns.patch
userns-allow-ptrace-from-non-init-user-namespaces.patch
userns-user-namespaces-convert-all-capable-checks-in-kernel-sysc.patch
userns-add-a-user-namespace-owner-of-ipc-ns.patch
userns-user-namespaces-convert-several-capable-calls.patch
userns-userns-check-user-namespace-for-task-file-uid-equivalence-checks.patch
userns-rename-is_owner_or_cap-to-inode_owner_or_capable.patch
--
To unsubscribe from this list: send the line "unsubscribe mm-commits" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
