2019-01-16 23:46:06 +08:00
|
|
|
// SPDX-License-Identifier: GPL-2.0
|
|
|
|
/*
|
|
|
|
* SafeSetID Linux Security Module
|
|
|
|
*
|
|
|
|
* Author: Micah Morton <mortonm@chromium.org>
|
|
|
|
*
|
|
|
|
* Copyright (C) 2018 The Chromium OS Authors.
|
|
|
|
*
|
|
|
|
* This program is free software; you can redistribute it and/or modify
|
|
|
|
* it under the terms of the GNU General Public License version 2, as
|
|
|
|
* published by the Free Software Foundation.
|
|
|
|
*
|
|
|
|
*/
|
|
|
|
|
|
|
|
#define pr_fmt(fmt) "SafeSetID: " fmt
|
|
|
|
|
|
|
|
#include <linux/lsm_hooks.h>
|
|
|
|
#include <linux/module.h>
|
|
|
|
#include <linux/ptrace.h>
|
|
|
|
#include <linux/sched/task_stack.h>
|
|
|
|
#include <linux/security.h>
|
2019-04-11 00:55:34 +08:00
|
|
|
#include "lsm.h"
|
2019-01-16 23:46:06 +08:00
|
|
|
|
|
|
|
/* Flag indicating whether initialization completed */
|
|
|
|
int safesetid_initialized;
|
|
|
|
|
2019-04-11 00:55:34 +08:00
|
|
|
#define NUM_BITS 8 /* 256 buckets in hash table */
|
2019-01-16 23:46:06 +08:00
|
|
|
|
|
|
|
static DEFINE_HASHTABLE(safesetid_whitelist_hashtable, NUM_BITS);
|
|
|
|
|
|
|
|
static DEFINE_SPINLOCK(safesetid_whitelist_hashtable_spinlock);
|
|
|
|
|
2019-04-11 00:55:34 +08:00
|
|
|
static enum sid_policy_type setuid_policy_lookup(kuid_t src, kuid_t dst)
|
2019-01-16 23:46:06 +08:00
|
|
|
{
|
|
|
|
struct entry *entry;
|
2019-04-11 00:55:34 +08:00
|
|
|
enum sid_policy_type result = SIDPOL_DEFAULT;
|
2019-01-16 23:46:06 +08:00
|
|
|
|
|
|
|
rcu_read_lock();
|
|
|
|
hash_for_each_possible_rcu(safesetid_whitelist_hashtable,
|
2019-04-11 00:55:34 +08:00
|
|
|
entry, next, __kuid_val(src)) {
|
|
|
|
if (!uid_eq(entry->src_uid, src))
|
|
|
|
continue;
|
|
|
|
if (uid_eq(entry->dst_uid, dst)) {
|
2019-01-16 23:46:06 +08:00
|
|
|
rcu_read_unlock();
|
2019-04-11 00:55:34 +08:00
|
|
|
return SIDPOL_ALLOWED;
|
2019-01-16 23:46:06 +08:00
|
|
|
}
|
2019-04-11 00:55:34 +08:00
|
|
|
result = SIDPOL_CONSTRAINED;
|
2019-01-16 23:46:06 +08:00
|
|
|
}
|
|
|
|
rcu_read_unlock();
|
2019-04-11 00:55:34 +08:00
|
|
|
return result;
|
2019-01-16 23:46:06 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
static int safesetid_security_capable(const struct cred *cred,
|
|
|
|
struct user_namespace *ns,
|
|
|
|
int cap,
|
|
|
|
unsigned int opts)
|
|
|
|
{
|
2019-04-11 00:55:41 +08:00
|
|
|
/* We're only interested in CAP_SETUID. */
|
|
|
|
if (cap != CAP_SETUID)
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If CAP_SETUID is currently used for a set*uid() syscall, we want to
|
|
|
|
* let it go through here; the real security check happens later, in the
|
|
|
|
* task_fix_setuid hook.
|
|
|
|
*/
|
|
|
|
if ((opts & CAP_OPT_INSETID) != 0)
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If no policy applies to this task, allow the use of CAP_SETUID for
|
|
|
|
* other purposes.
|
|
|
|
*/
|
|
|
|
if (setuid_policy_lookup(cred->uid, INVALID_UID) == SIDPOL_DEFAULT)
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Reject use of CAP_SETUID for functionality other than calling
|
|
|
|
* set*uid() (e.g. setting up userns uid mappings).
|
|
|
|
*/
|
|
|
|
pr_warn("Operation requires CAP_SETUID, which is not available to UID %u for operations besides approved set*uid transitions\n",
|
|
|
|
__kuid_val(cred->uid));
|
|
|
|
return -1;
|
2019-01-16 23:46:06 +08:00
|
|
|
}
|
|
|
|
|
LSM: SafeSetID: fix check for setresuid(new1, new2, new3)
With the old code, when a process with the (real,effective,saved) UID set
(1,1,1) calls setresuid(2,3,4), safesetid_task_fix_setuid() only checks
whether the transition 1->2 is permitted; the transitions 1->3 and 1->4 are
not checked. Fix this.
This is also a good opportunity to refactor safesetid_task_fix_setuid() to
be less verbose - having one branch per set*uid() syscall is unnecessary.
Note that this slightly changes semantics: The UID transition check for
UIDs that were not in the old cred struct is now always performed against
the policy of the RUID. I think that's more consistent anyway, since the
RUID is also the one that decides whether any policy is enforced at all.
Signed-off-by: Jann Horn <jannh@google.com>
Signed-off-by: Micah Morton <mortonm@chromium.org>
2019-04-11 00:55:19 +08:00
|
|
|
/*
|
|
|
|
* Check whether a caller with old credentials @old is allowed to switch to
|
|
|
|
* credentials that contain @new_uid.
|
|
|
|
*/
|
|
|
|
static bool uid_permitted_for_cred(const struct cred *old, kuid_t new_uid)
|
2019-01-16 23:46:06 +08:00
|
|
|
{
|
LSM: SafeSetID: fix check for setresuid(new1, new2, new3)
With the old code, when a process with the (real,effective,saved) UID set
(1,1,1) calls setresuid(2,3,4), safesetid_task_fix_setuid() only checks
whether the transition 1->2 is permitted; the transitions 1->3 and 1->4 are
not checked. Fix this.
This is also a good opportunity to refactor safesetid_task_fix_setuid() to
be less verbose - having one branch per set*uid() syscall is unnecessary.
Note that this slightly changes semantics: The UID transition check for
UIDs that were not in the old cred struct is now always performed against
the policy of the RUID. I think that's more consistent anyway, since the
RUID is also the one that decides whether any policy is enforced at all.
Signed-off-by: Jann Horn <jannh@google.com>
Signed-off-by: Micah Morton <mortonm@chromium.org>
2019-04-11 00:55:19 +08:00
|
|
|
bool permitted;
|
|
|
|
|
|
|
|
/* If our old creds already had this UID in it, it's fine. */
|
|
|
|
if (uid_eq(new_uid, old->uid) || uid_eq(new_uid, old->euid) ||
|
|
|
|
uid_eq(new_uid, old->suid))
|
|
|
|
return true;
|
|
|
|
|
2019-01-16 23:46:06 +08:00
|
|
|
/*
|
LSM: SafeSetID: fix check for setresuid(new1, new2, new3)
With the old code, when a process with the (real,effective,saved) UID set
(1,1,1) calls setresuid(2,3,4), safesetid_task_fix_setuid() only checks
whether the transition 1->2 is permitted; the transitions 1->3 and 1->4 are
not checked. Fix this.
This is also a good opportunity to refactor safesetid_task_fix_setuid() to
be less verbose - having one branch per set*uid() syscall is unnecessary.
Note that this slightly changes semantics: The UID transition check for
UIDs that were not in the old cred struct is now always performed against
the policy of the RUID. I think that's more consistent anyway, since the
RUID is also the one that decides whether any policy is enforced at all.
Signed-off-by: Jann Horn <jannh@google.com>
Signed-off-by: Micah Morton <mortonm@chromium.org>
2019-04-11 00:55:19 +08:00
|
|
|
* Transitions to new UIDs require a check against the policy of the old
|
|
|
|
* RUID.
|
2019-01-16 23:46:06 +08:00
|
|
|
*/
|
2019-04-11 00:55:34 +08:00
|
|
|
permitted =
|
|
|
|
setuid_policy_lookup(old->uid, new_uid) != SIDPOL_CONSTRAINED;
|
LSM: SafeSetID: fix check for setresuid(new1, new2, new3)
With the old code, when a process with the (real,effective,saved) UID set
(1,1,1) calls setresuid(2,3,4), safesetid_task_fix_setuid() only checks
whether the transition 1->2 is permitted; the transitions 1->3 and 1->4 are
not checked. Fix this.
This is also a good opportunity to refactor safesetid_task_fix_setuid() to
be less verbose - having one branch per set*uid() syscall is unnecessary.
Note that this slightly changes semantics: The UID transition check for
UIDs that were not in the old cred struct is now always performed against
the policy of the RUID. I think that's more consistent anyway, since the
RUID is also the one that decides whether any policy is enforced at all.
Signed-off-by: Jann Horn <jannh@google.com>
Signed-off-by: Micah Morton <mortonm@chromium.org>
2019-04-11 00:55:19 +08:00
|
|
|
if (!permitted) {
|
|
|
|
pr_warn("UID transition ((%d,%d,%d) -> %d) blocked\n",
|
|
|
|
__kuid_val(old->uid), __kuid_val(old->euid),
|
|
|
|
__kuid_val(old->suid), __kuid_val(new_uid));
|
|
|
|
}
|
|
|
|
return permitted;
|
2019-01-16 23:46:06 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Check whether there is either an exception for user under old cred struct to
|
|
|
|
* set*uid to user under new cred struct, or the UID transition is allowed (by
|
|
|
|
* Linux set*uid rules) even without CAP_SETUID.
|
|
|
|
*/
|
|
|
|
static int safesetid_task_fix_setuid(struct cred *new,
|
|
|
|
const struct cred *old,
|
|
|
|
int flags)
|
|
|
|
{
|
|
|
|
|
LSM: SafeSetID: fix check for setresuid(new1, new2, new3)
With the old code, when a process with the (real,effective,saved) UID set
(1,1,1) calls setresuid(2,3,4), safesetid_task_fix_setuid() only checks
whether the transition 1->2 is permitted; the transitions 1->3 and 1->4 are
not checked. Fix this.
This is also a good opportunity to refactor safesetid_task_fix_setuid() to
be less verbose - having one branch per set*uid() syscall is unnecessary.
Note that this slightly changes semantics: The UID transition check for
UIDs that were not in the old cred struct is now always performed against
the policy of the RUID. I think that's more consistent anyway, since the
RUID is also the one that decides whether any policy is enforced at all.
Signed-off-by: Jann Horn <jannh@google.com>
Signed-off-by: Micah Morton <mortonm@chromium.org>
2019-04-11 00:55:19 +08:00
|
|
|
/* Do nothing if there are no setuid restrictions for our old RUID. */
|
2019-04-11 00:55:34 +08:00
|
|
|
if (setuid_policy_lookup(old->uid, INVALID_UID) == SIDPOL_DEFAULT)
|
2019-01-16 23:46:06 +08:00
|
|
|
return 0;
|
|
|
|
|
LSM: SafeSetID: fix check for setresuid(new1, new2, new3)
With the old code, when a process with the (real,effective,saved) UID set
(1,1,1) calls setresuid(2,3,4), safesetid_task_fix_setuid() only checks
whether the transition 1->2 is permitted; the transitions 1->3 and 1->4 are
not checked. Fix this.
This is also a good opportunity to refactor safesetid_task_fix_setuid() to
be less verbose - having one branch per set*uid() syscall is unnecessary.
Note that this slightly changes semantics: The UID transition check for
UIDs that were not in the old cred struct is now always performed against
the policy of the RUID. I think that's more consistent anyway, since the
RUID is also the one that decides whether any policy is enforced at all.
Signed-off-by: Jann Horn <jannh@google.com>
Signed-off-by: Micah Morton <mortonm@chromium.org>
2019-04-11 00:55:19 +08:00
|
|
|
if (uid_permitted_for_cred(old, new->uid) &&
|
|
|
|
uid_permitted_for_cred(old, new->euid) &&
|
|
|
|
uid_permitted_for_cred(old, new->suid) &&
|
|
|
|
uid_permitted_for_cred(old, new->fsuid))
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Kill this process to avoid potential security vulnerabilities
|
|
|
|
* that could arise from a missing whitelist entry preventing a
|
|
|
|
* privileged process from dropping to a lesser-privileged one.
|
|
|
|
*/
|
|
|
|
force_sig(SIGKILL);
|
|
|
|
return -EACCES;
|
2019-01-16 23:46:06 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
int add_safesetid_whitelist_entry(kuid_t parent, kuid_t child)
|
|
|
|
{
|
|
|
|
struct entry *new;
|
|
|
|
|
|
|
|
/* Return if entry already exists */
|
2019-04-11 00:55:34 +08:00
|
|
|
if (setuid_policy_lookup(parent, child) == SIDPOL_ALLOWED)
|
2019-01-16 23:46:06 +08:00
|
|
|
return 0;
|
|
|
|
|
|
|
|
new = kzalloc(sizeof(struct entry), GFP_KERNEL);
|
|
|
|
if (!new)
|
|
|
|
return -ENOMEM;
|
2019-04-11 00:55:34 +08:00
|
|
|
new->src_uid = parent;
|
|
|
|
new->dst_uid = child;
|
2019-01-16 23:46:06 +08:00
|
|
|
spin_lock(&safesetid_whitelist_hashtable_spinlock);
|
|
|
|
hash_add_rcu(safesetid_whitelist_hashtable,
|
|
|
|
&new->next,
|
|
|
|
__kuid_val(parent));
|
|
|
|
spin_unlock(&safesetid_whitelist_hashtable_spinlock);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
void flush_safesetid_whitelist_entries(void)
|
|
|
|
{
|
|
|
|
struct entry *entry;
|
|
|
|
struct hlist_node *hlist_node;
|
|
|
|
unsigned int bkt_loop_cursor;
|
|
|
|
HLIST_HEAD(free_list);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Could probably use hash_for_each_rcu here instead, but this should
|
|
|
|
* be fine as well.
|
|
|
|
*/
|
|
|
|
spin_lock(&safesetid_whitelist_hashtable_spinlock);
|
|
|
|
hash_for_each_safe(safesetid_whitelist_hashtable, bkt_loop_cursor,
|
|
|
|
hlist_node, entry, next) {
|
|
|
|
hash_del_rcu(&entry->next);
|
|
|
|
hlist_add_head(&entry->dlist, &free_list);
|
|
|
|
}
|
|
|
|
spin_unlock(&safesetid_whitelist_hashtable_spinlock);
|
|
|
|
synchronize_rcu();
|
|
|
|
hlist_for_each_entry_safe(entry, hlist_node, &free_list, dlist) {
|
|
|
|
hlist_del(&entry->dlist);
|
|
|
|
kfree(entry);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static struct security_hook_list safesetid_security_hooks[] = {
|
|
|
|
LSM_HOOK_INIT(task_fix_setuid, safesetid_task_fix_setuid),
|
|
|
|
LSM_HOOK_INIT(capable, safesetid_security_capable)
|
|
|
|
};
|
|
|
|
|
|
|
|
static int __init safesetid_security_init(void)
|
|
|
|
{
|
|
|
|
security_add_hooks(safesetid_security_hooks,
|
|
|
|
ARRAY_SIZE(safesetid_security_hooks), "safesetid");
|
|
|
|
|
|
|
|
/* Report that SafeSetID successfully initialized */
|
|
|
|
safesetid_initialized = 1;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
DEFINE_LSM(safesetid_security_init) = {
|
|
|
|
.init = safesetid_security_init,
|
2019-01-29 04:30:56 +08:00
|
|
|
.name = "safesetid",
|
2019-01-16 23:46:06 +08:00
|
|
|
};
|