#include <linux/bitops.h>
#include <linux/key.h>
#include <linux/interrupt.h>
+#include <linux/module.h>
+#include <linux/user_namespace.h>
+#include "cred-internals.h"
+
+struct user_namespace init_user_ns = {
+ .kref = {
+ .refcount = ATOMIC_INIT(2),
+ },
+ .creator = &root_user,
+};
+EXPORT_SYMBOL_GPL(init_user_ns);
/*
* UID task count cache, to get fast user lookup in "alloc_uid"
* when changing user ID's (ie setuid() and friends).
*/
-#define UIDHASH_BITS (CONFIG_BASE_SMALL ? 3 : 8)
-#define UIDHASH_SZ (1 << UIDHASH_BITS)
#define UIDHASH_MASK (UIDHASH_SZ - 1)
#define __uidhashfn(uid) (((uid >> UIDHASH_BITS) + uid) & UIDHASH_MASK)
-#define uidhashentry(uid) (uidhash_table + __uidhashfn((uid)))
+#define uidhashentry(ns, uid) ((ns)->uidhash_table + __uidhashfn((uid)))
-static kmem_cache_t *uid_cachep;
-static struct list_head uidhash_table[UIDHASH_SZ];
+static struct kmem_cache *uid_cachep;
/*
* The uidhash_lock is mostly taken from process context, but it is
*/
static DEFINE_SPINLOCK(uidhash_lock);
+/* root_user.__count is 2, 1 for init task cred, 1 for init_user_ns->creator */
struct user_struct root_user = {
- .__count = ATOMIC_INIT(1),
+ .__count = ATOMIC_INIT(2),
.processes = ATOMIC_INIT(1),
.files = ATOMIC_INIT(0),
.sigpending = ATOMIC_INIT(0),
- .mq_bytes = 0,
.locked_shm = 0,
-#ifdef CONFIG_KEYS
- .uid_keyring = &root_user_keyring,
- .session_keyring = &root_session_keyring,
+ .user_ns = &init_user_ns,
+#ifdef CONFIG_USER_SCHED
+ .tg = &init_task_group,
#endif
};
/*
* These routines must be called with the uidhash spinlock held!
*/
-static inline void uid_hash_insert(struct user_struct *up, struct list_head *hashent)
+static void uid_hash_insert(struct user_struct *up, struct hlist_head *hashent)
{
- list_add(&up->uidhash_list, hashent);
+ hlist_add_head(&up->uidhash_node, hashent);
}
-static inline void uid_hash_remove(struct user_struct *up)
+static void uid_hash_remove(struct user_struct *up)
+{
+ hlist_del_init(&up->uidhash_node);
+ put_user_ns(up->user_ns);
+}
+
+#ifdef CONFIG_USER_SCHED
+
+static void sched_destroy_user(struct user_struct *up)
{
- list_del(&up->uidhash_list);
+ sched_destroy_group(up->tg);
}
-static inline struct user_struct *uid_hash_find(uid_t uid, struct list_head *hashent)
+static int sched_create_user(struct user_struct *up)
{
- struct list_head *up;
+ int rc = 0;
+
+ up->tg = sched_create_group(&root_task_group);
+ if (IS_ERR(up->tg))
+ rc = -ENOMEM;
+
+ set_tg_uid(up);
+
+ return rc;
+}
+
+#else /* CONFIG_USER_SCHED */
- list_for_each(up, hashent) {
- struct user_struct *user;
+static void sched_destroy_user(struct user_struct *up) { }
+static int sched_create_user(struct user_struct *up) { return 0; }
- user = list_entry(up, struct user_struct, uidhash_list);
+#endif /* CONFIG_USER_SCHED */
+
+#if defined(CONFIG_USER_SCHED) && defined(CONFIG_SYSFS)
+
+static struct user_struct *uid_hash_find(uid_t uid, struct hlist_head *hashent)
+{
+ struct user_struct *user;
+ struct hlist_node *h;
+
+ hlist_for_each_entry(user, h, hashent, uidhash_node) {
+ if (user->uid == uid) {
+ /* possibly resurrect an "almost deleted" object */
+ if (atomic_inc_return(&user->__count) == 1)
+ cancel_delayed_work(&user->work);
+ return user;
+ }
+ }
+
+ return NULL;
+}
- if(user->uid == uid) {
+static struct kset *uids_kset; /* represents the /sys/kernel/uids/ directory */
+static DEFINE_MUTEX(uids_mutex);
+
+static inline void uids_mutex_lock(void)
+{
+ mutex_lock(&uids_mutex);
+}
+
+static inline void uids_mutex_unlock(void)
+{
+ mutex_unlock(&uids_mutex);
+}
+
+/* uid directory attributes */
+#ifdef CONFIG_FAIR_GROUP_SCHED
+static ssize_t cpu_shares_show(struct kobject *kobj,
+ struct kobj_attribute *attr,
+ char *buf)
+{
+ struct user_struct *up = container_of(kobj, struct user_struct, kobj);
+
+ return sprintf(buf, "%lu\n", sched_group_shares(up->tg));
+}
+
+static ssize_t cpu_shares_store(struct kobject *kobj,
+ struct kobj_attribute *attr,
+ const char *buf, size_t size)
+{
+ struct user_struct *up = container_of(kobj, struct user_struct, kobj);
+ unsigned long shares;
+ int rc;
+
+ sscanf(buf, "%lu", &shares);
+
+ rc = sched_group_set_shares(up->tg, shares);
+
+ return (rc ? rc : size);
+}
+
+static struct kobj_attribute cpu_share_attr =
+ __ATTR(cpu_share, 0644, cpu_shares_show, cpu_shares_store);
+#endif
+
+#ifdef CONFIG_RT_GROUP_SCHED
+static ssize_t cpu_rt_runtime_show(struct kobject *kobj,
+ struct kobj_attribute *attr,
+ char *buf)
+{
+ struct user_struct *up = container_of(kobj, struct user_struct, kobj);
+
+ return sprintf(buf, "%ld\n", sched_group_rt_runtime(up->tg));
+}
+
+static ssize_t cpu_rt_runtime_store(struct kobject *kobj,
+ struct kobj_attribute *attr,
+ const char *buf, size_t size)
+{
+ struct user_struct *up = container_of(kobj, struct user_struct, kobj);
+ unsigned long rt_runtime;
+ int rc;
+
+ sscanf(buf, "%ld", &rt_runtime);
+
+ rc = sched_group_set_rt_runtime(up->tg, rt_runtime);
+
+ return (rc ? rc : size);
+}
+
+static struct kobj_attribute cpu_rt_runtime_attr =
+ __ATTR(cpu_rt_runtime, 0644, cpu_rt_runtime_show, cpu_rt_runtime_store);
+
+static ssize_t cpu_rt_period_show(struct kobject *kobj,
+ struct kobj_attribute *attr,
+ char *buf)
+{
+ struct user_struct *up = container_of(kobj, struct user_struct, kobj);
+
+ return sprintf(buf, "%lu\n", sched_group_rt_period(up->tg));
+}
+
+static ssize_t cpu_rt_period_store(struct kobject *kobj,
+ struct kobj_attribute *attr,
+ const char *buf, size_t size)
+{
+ struct user_struct *up = container_of(kobj, struct user_struct, kobj);
+ unsigned long rt_period;
+ int rc;
+
+ sscanf(buf, "%lu", &rt_period);
+
+ rc = sched_group_set_rt_period(up->tg, rt_period);
+
+ return (rc ? rc : size);
+}
+
+static struct kobj_attribute cpu_rt_period_attr =
+ __ATTR(cpu_rt_period, 0644, cpu_rt_period_show, cpu_rt_period_store);
+#endif
+
+/* default attributes per uid directory */
+static struct attribute *uids_attributes[] = {
+#ifdef CONFIG_FAIR_GROUP_SCHED
+ &cpu_share_attr.attr,
+#endif
+#ifdef CONFIG_RT_GROUP_SCHED
+ &cpu_rt_runtime_attr.attr,
+ &cpu_rt_period_attr.attr,
+#endif
+ NULL
+};
+
+/* the lifetime of user_struct is not managed by the core (now) */
+static void uids_release(struct kobject *kobj)
+{
+ return;
+}
+
+static struct kobj_type uids_ktype = {
+ .sysfs_ops = &kobj_sysfs_ops,
+ .default_attrs = uids_attributes,
+ .release = uids_release,
+};
+
+/*
+ * Create /sys/kernel/uids/<uid>/cpu_share file for this user
+ * We do not create this file for users in a user namespace (until
+ * sysfs tagging is implemented).
+ *
+ * See Documentation/scheduler/sched-design-CFS.txt for ramifications.
+ */
+static int uids_user_create(struct user_struct *up)
+{
+ struct kobject *kobj = &up->kobj;
+ int error;
+
+ memset(kobj, 0, sizeof(struct kobject));
+ if (up->user_ns != &init_user_ns)
+ return 0;
+ kobj->kset = uids_kset;
+ error = kobject_init_and_add(kobj, &uids_ktype, NULL, "%d", up->uid);
+ if (error) {
+ kobject_put(kobj);
+ goto done;
+ }
+
+ kobject_uevent(kobj, KOBJ_ADD);
+done:
+ return error;
+}
+
+/* create these entries in sysfs:
+ * "/sys/kernel/uids" directory
+ * "/sys/kernel/uids/0" directory (for root user)
+ * "/sys/kernel/uids/0/cpu_share" file (for root user)
+ */
+int __init uids_sysfs_init(void)
+{
+ uids_kset = kset_create_and_add("uids", NULL, kernel_kobj);
+ if (!uids_kset)
+ return -ENOMEM;
+
+ return uids_user_create(&root_user);
+}
+
+/* delayed work function to remove sysfs directory for a user and free up
+ * corresponding structures.
+ */
+static void cleanup_user_struct(struct work_struct *w)
+{
+ struct user_struct *up = container_of(w, struct user_struct, work.work);
+ unsigned long flags;
+ int remove_user = 0;
+
+ /* Make uid_hash_remove() + sysfs_remove_file() + kobject_del()
+ * atomic.
+ */
+ uids_mutex_lock();
+
+ spin_lock_irqsave(&uidhash_lock, flags);
+ if (atomic_read(&up->__count) == 0) {
+ uid_hash_remove(up);
+ remove_user = 1;
+ }
+ spin_unlock_irqrestore(&uidhash_lock, flags);
+
+ if (!remove_user)
+ goto done;
+
+ if (up->user_ns == &init_user_ns) {
+ kobject_uevent(&up->kobj, KOBJ_REMOVE);
+ kobject_del(&up->kobj);
+ kobject_put(&up->kobj);
+ }
+
+ sched_destroy_user(up);
+ key_put(up->uid_keyring);
+ key_put(up->session_keyring);
+ kmem_cache_free(uid_cachep, up);
+
+done:
+ uids_mutex_unlock();
+}
+
+/* IRQs are disabled and uidhash_lock is held upon function entry.
+ * IRQ state (as stored in flags) is restored and uidhash_lock released
+ * upon function exit.
+ */
+static void free_user(struct user_struct *up, unsigned long flags)
+{
+ INIT_DELAYED_WORK(&up->work, cleanup_user_struct);
+ schedule_delayed_work(&up->work, msecs_to_jiffies(1000));
+ spin_unlock_irqrestore(&uidhash_lock, flags);
+}
+
+#else /* CONFIG_USER_SCHED && CONFIG_SYSFS */
+
+static struct user_struct *uid_hash_find(uid_t uid, struct hlist_head *hashent)
+{
+ struct user_struct *user;
+ struct hlist_node *h;
+
+ hlist_for_each_entry(user, h, hashent, uidhash_node) {
+ if (user->uid == uid) {
atomic_inc(&user->__count);
return user;
}
return NULL;
}
+int uids_sysfs_init(void) { return 0; }
+static inline int uids_user_create(struct user_struct *up) { return 0; }
+static inline void uids_mutex_lock(void) { }
+static inline void uids_mutex_unlock(void) { }
+
+/* IRQs are disabled and uidhash_lock is held upon function entry.
+ * IRQ state (as stored in flags) is restored and uidhash_lock released
+ * upon function exit.
+ */
+static void free_user(struct user_struct *up, unsigned long flags)
+{
+ uid_hash_remove(up);
+ spin_unlock_irqrestore(&uidhash_lock, flags);
+ sched_destroy_user(up);
+ key_put(up->uid_keyring);
+ key_put(up->session_keyring);
+ kmem_cache_free(uid_cachep, up);
+}
+
+#endif
+
+#if defined(CONFIG_RT_GROUP_SCHED) && defined(CONFIG_USER_SCHED)
+/*
+ * We need to check if a setuid can take place. This function should be called
+ * before successfully completing the setuid.
+ */
+int task_can_switch_user(struct user_struct *up, struct task_struct *tsk)
+{
+
+ return sched_rt_can_attach(up->tg, tsk);
+
+}
+#else
+int task_can_switch_user(struct user_struct *up, struct task_struct *tsk)
+{
+ return 1;
+}
+#endif
+
/*
* Locate the user_struct for the passed UID. If found, take a ref on it. The
* caller must undo that ref with free_uid().
{
struct user_struct *ret;
unsigned long flags;
+ struct user_namespace *ns = current_user_ns();
spin_lock_irqsave(&uidhash_lock, flags);
- ret = uid_hash_find(uid, uidhashentry(uid));
+ ret = uid_hash_find(uid, uidhashentry(ns, uid));
spin_unlock_irqrestore(&uidhash_lock, flags);
return ret;
}
return;
local_irq_save(flags);
- if (atomic_dec_and_lock(&up->__count, &uidhash_lock)) {
- uid_hash_remove(up);
- spin_unlock_irqrestore(&uidhash_lock, flags);
- key_put(up->uid_keyring);
- key_put(up->session_keyring);
- kmem_cache_free(uid_cachep, up);
- } else {
+ if (atomic_dec_and_lock(&up->__count, &uidhash_lock))
+ free_user(up, flags);
+ else
local_irq_restore(flags);
- }
}
-struct user_struct * alloc_uid(uid_t uid)
+struct user_struct *alloc_uid(struct user_namespace *ns, uid_t uid)
{
- struct list_head *hashent = uidhashentry(uid);
- struct user_struct *up;
+ struct hlist_head *hashent = uidhashentry(ns, uid);
+ struct user_struct *up, *new;
+
+ /* Make uid_hash_find() + uids_user_create() + uid_hash_insert()
+ * atomic.
+ */
+ uids_mutex_lock();
spin_lock_irq(&uidhash_lock);
up = uid_hash_find(uid, hashent);
spin_unlock_irq(&uidhash_lock);
if (!up) {
- struct user_struct *new;
-
- new = kmem_cache_alloc(uid_cachep, SLAB_KERNEL);
+ new = kmem_cache_zalloc(uid_cachep, GFP_KERNEL);
if (!new)
- return NULL;
+ goto out_unlock;
+
new->uid = uid;
atomic_set(&new->__count, 1);
- atomic_set(&new->processes, 0);
- atomic_set(&new->files, 0);
- atomic_set(&new->sigpending, 0);
-#ifdef CONFIG_INOTIFY_USER
- atomic_set(&new->inotify_watches, 0);
- atomic_set(&new->inotify_devs, 0);
-#endif
- new->mq_bytes = 0;
- new->locked_shm = 0;
+ if (sched_create_user(new) < 0)
+ goto out_free_user;
- if (alloc_uid_keyring(new, current) < 0) {
- kmem_cache_free(uid_cachep, new);
- return NULL;
- }
+ new->user_ns = get_user_ns(ns);
+
+ if (uids_user_create(new))
+ goto out_destoy_sched;
/*
* Before adding this, check whether we raced
spin_lock_irq(&uidhash_lock);
up = uid_hash_find(uid, hashent);
if (up) {
+ /* This case is not possible when CONFIG_USER_SCHED
+ * is defined, since we serialize alloc_uid() using
+ * uids_mutex. Hence no need to call
+ * sched_destroy_user() or remove_user_sysfs_dir().
+ */
key_put(new->uid_keyring);
key_put(new->session_keyring);
kmem_cache_free(uid_cachep, new);
up = new;
}
spin_unlock_irq(&uidhash_lock);
-
}
- return up;
-}
-void switch_uid(struct user_struct *new_user)
-{
- struct user_struct *old_user;
+ uids_mutex_unlock();
- /* What if a process setreuid()'s and this brings the
- * new uid over his NPROC rlimit? We can check this now
- * cheaply with the new uid cache, so if it matters
- * we should be checking for it. -DaveM
- */
- old_user = current->user;
- atomic_inc(&new_user->processes);
- atomic_dec(&old_user->processes);
- switch_uid_keyring(new_user);
- current->user = new_user;
-
- /*
- * We need to synchronize with __sigqueue_alloc()
- * doing a get_uid(p->user).. If that saw the old
- * user value, we need to wait until it has exited
- * its critical region before we can free the old
- * structure.
- */
- smp_mb();
- spin_unlock_wait(¤t->sighand->siglock);
+ return up;
- free_uid(old_user);
- suid_keys(current);
+out_destoy_sched:
+ sched_destroy_user(new);
+ put_user_ns(new->user_ns);
+out_free_user:
+ kmem_cache_free(uid_cachep, new);
+out_unlock:
+ uids_mutex_unlock();
+ return NULL;
}
-
static int __init uid_cache_init(void)
{
int n;
uid_cachep = kmem_cache_create("uid_cache", sizeof(struct user_struct),
- 0, SLAB_HWCACHE_ALIGN|SLAB_PANIC, NULL, NULL);
+ 0, SLAB_HWCACHE_ALIGN|SLAB_PANIC, NULL);
for(n = 0; n < UIDHASH_SZ; ++n)
- INIT_LIST_HEAD(uidhash_table + n);
+ INIT_HLIST_HEAD(init_user_ns.uidhash_table + n);
/* Insert the root user immediately (init already runs as root) */
spin_lock_irq(&uidhash_lock);
- uid_hash_insert(&root_user, uidhashentry(0));
+ uid_hash_insert(&root_user, uidhashentry(&init_user_ns, 0));
spin_unlock_irq(&uidhash_lock);
return 0;
Code Review / linux-2.6.git / blobdiff