summaryrefslogtreecommitdiff
path: root/kernel/cgroup/cgroup-v1.c
diff options
context:
space:
mode:
Diffstat (limited to 'kernel/cgroup/cgroup-v1.c')
-rw-r--r--kernel/cgroup/cgroup-v1.c206
1 files changed, 141 insertions, 65 deletions
diff --git a/kernel/cgroup/cgroup-v1.c b/kernel/cgroup/cgroup-v1.c
index a5751784ad74..e28d5f0d20ed 100644
--- a/kernel/cgroup/cgroup-v1.c
+++ b/kernel/cgroup/cgroup-v1.c
@@ -46,24 +46,29 @@ bool cgroup1_ssid_disabled(int ssid)
return cgroup_no_v1_mask & (1 << ssid);
}
+static bool cgroup1_subsys_absent(struct cgroup_subsys *ss)
+{
+ /* Check also dfl_cftypes for file-less controllers, i.e. perf_event */
+ return ss->legacy_cftypes == NULL && ss->dfl_cftypes;
+}
+
/**
* cgroup_attach_task_all - attach task 'tsk' to all cgroups of task 'from'
* @from: attach to all cgroups of a given task
* @tsk: the task to be attached
+ *
+ * Return: %0 on success or a negative errno code on failure
*/
int cgroup_attach_task_all(struct task_struct *from, struct task_struct *tsk)
{
struct cgroup_root *root;
int retval = 0;
- mutex_lock(&cgroup_mutex);
- percpu_down_write(&cgroup_threadgroup_rwsem);
+ cgroup_lock();
+ cgroup_attach_lock(true);
for_each_root(root) {
struct cgroup *from_cgrp;
- if (root == &cgrp_dfl_root)
- continue;
-
spin_lock_irq(&css_set_lock);
from_cgrp = task_cgroup_from_root(from, root);
spin_unlock_irq(&css_set_lock);
@@ -72,15 +77,15 @@ int cgroup_attach_task_all(struct task_struct *from, struct task_struct *tsk)
if (retval)
break;
}
- percpu_up_write(&cgroup_threadgroup_rwsem);
- mutex_unlock(&cgroup_mutex);
+ cgroup_attach_unlock(true);
+ cgroup_unlock();
return retval;
}
EXPORT_SYMBOL_GPL(cgroup_attach_task_all);
/**
- * cgroup_trasnsfer_tasks - move tasks from one cgroup to another
+ * cgroup_transfer_tasks - move tasks from one cgroup to another
* @to: cgroup to which the tasks will be moved
* @from: cgroup in which the tasks currently reside
*
@@ -89,6 +94,8 @@ EXPORT_SYMBOL_GPL(cgroup_attach_task_all);
* is guaranteed to be either visible in the source cgroup after the
* parent's migration is complete or put into the target cgroup. No task
* can slip out of migration through forking.
+ *
+ * Return: %0 on success or a negative errno code on failure
*/
int cgroup_transfer_tasks(struct cgroup *to, struct cgroup *from)
{
@@ -105,9 +112,9 @@ int cgroup_transfer_tasks(struct cgroup *to, struct cgroup *from)
if (ret)
return ret;
- mutex_lock(&cgroup_mutex);
+ cgroup_lock();
- percpu_down_write(&cgroup_threadgroup_rwsem);
+ cgroup_attach_lock(true);
/* all tasks in @from are being moved, all csets are source */
spin_lock_irq(&css_set_lock);
@@ -143,8 +150,8 @@ int cgroup_transfer_tasks(struct cgroup *to, struct cgroup *from)
} while (task && !ret);
out_err:
cgroup_migrate_finish(&mgctx);
- percpu_up_write(&cgroup_threadgroup_rwsem);
- mutex_unlock(&cgroup_mutex);
+ cgroup_attach_unlock(true);
+ cgroup_unlock();
return ret;
}
@@ -359,10 +366,9 @@ static int pidlist_array_load(struct cgroup *cgrp, enum cgroup_filetype type,
}
css_task_iter_end(&it);
length = n;
- /* now sort & (if procs) strip out duplicates */
+ /* now sort & strip out duplicates (tgids or recycled thread PIDs) */
sort(array, length, sizeof(pid_t), cmppid, NULL);
- if (type == CGROUP_FILE_PROCS)
- length = pidlist_uniq(array, length);
+ length = pidlist_uniq(array, length);
l = cgroup_pidlist_find_create(cgrp, type);
if (!l) {
@@ -393,6 +399,7 @@ static void *cgroup_pidlist_start(struct seq_file *s, loff_t *pos)
* next pid to display, if any
*/
struct kernfs_open_file *of = s->private;
+ struct cgroup_file_ctx *ctx = of->priv;
struct cgroup *cgrp = seq_css(s)->cgroup;
struct cgroup_pidlist *l;
enum cgroup_filetype type = seq_cft(s)->private;
@@ -402,25 +409,24 @@ static void *cgroup_pidlist_start(struct seq_file *s, loff_t *pos)
mutex_lock(&cgrp->pidlist_mutex);
/*
- * !NULL @of->priv indicates that this isn't the first start()
- * after open. If the matching pidlist is around, we can use that.
- * Look for it. Note that @of->priv can't be used directly. It
- * could already have been destroyed.
+ * !NULL @ctx->procs1.pidlist indicates that this isn't the first
+ * start() after open. If the matching pidlist is around, we can use
+ * that. Look for it. Note that @ctx->procs1.pidlist can't be used
+ * directly. It could already have been destroyed.
*/
- if (of->priv)
- of->priv = cgroup_pidlist_find(cgrp, type);
+ if (ctx->procs1.pidlist)
+ ctx->procs1.pidlist = cgroup_pidlist_find(cgrp, type);
/*
* Either this is the first start() after open or the matching
* pidlist has been destroyed inbetween. Create a new one.
*/
- if (!of->priv) {
- ret = pidlist_array_load(cgrp, type,
- (struct cgroup_pidlist **)&of->priv);
+ if (!ctx->procs1.pidlist) {
+ ret = pidlist_array_load(cgrp, type, &ctx->procs1.pidlist);
if (ret)
return ERR_PTR(ret);
}
- l = of->priv;
+ l = ctx->procs1.pidlist;
if (pid) {
int end = l->length;
@@ -430,7 +436,7 @@ static void *cgroup_pidlist_start(struct seq_file *s, loff_t *pos)
if (l->list[mid] == pid) {
index = mid;
break;
- } else if (l->list[mid] <= pid)
+ } else if (l->list[mid] < pid)
index = mid + 1;
else
end = mid;
@@ -448,7 +454,8 @@ static void *cgroup_pidlist_start(struct seq_file *s, loff_t *pos)
static void cgroup_pidlist_stop(struct seq_file *s, void *v)
{
struct kernfs_open_file *of = s->private;
- struct cgroup_pidlist *l = of->priv;
+ struct cgroup_file_ctx *ctx = of->priv;
+ struct cgroup_pidlist *l = ctx->procs1.pidlist;
if (l)
mod_delayed_work(cgroup_pidlist_destroy_wq, &l->destroy_dwork,
@@ -459,7 +466,8 @@ static void cgroup_pidlist_stop(struct seq_file *s, void *v)
static void *cgroup_pidlist_next(struct seq_file *s, void *v, loff_t *pos)
{
struct kernfs_open_file *of = s->private;
- struct cgroup_pidlist *l = of->priv;
+ struct cgroup_file_ctx *ctx = of->priv;
+ struct cgroup_pidlist *l = ctx->procs1.pidlist;
pid_t *p = v;
pid_t *end = l->list + l->length;
/*
@@ -503,10 +511,11 @@ static ssize_t __cgroup1_procs_write(struct kernfs_open_file *of,
goto out_unlock;
/*
- * Even if we're attaching all tasks in the thread group, we only
- * need to check permissions on one of them.
+ * Even if we're attaching all tasks in the thread group, we only need
+ * to check permissions on one of them. Check permissions using the
+ * credentials from file open to protect against inherited fd attacks.
*/
- cred = current_cred();
+ cred = of->file->f_cred;
tcred = get_task_cred(task);
if (!uid_eq(cred->euid, GLOBAL_ROOT_UID) &&
!uid_eq(cred->euid, tcred->uid) &&
@@ -542,14 +551,24 @@ static ssize_t cgroup_release_agent_write(struct kernfs_open_file *of,
char *buf, size_t nbytes, loff_t off)
{
struct cgroup *cgrp;
+ struct cgroup_file_ctx *ctx;
BUILD_BUG_ON(sizeof(cgrp->root->release_agent_path) < PATH_MAX);
+ /*
+ * Release agent gets called with all capabilities,
+ * require capabilities to set release agent.
+ */
+ ctx = of->priv;
+ if ((ctx->ns->user_ns != &init_user_ns) ||
+ !file_ns_capable(of->file, &init_user_ns, CAP_SYS_ADMIN))
+ return -EPERM;
+
cgrp = cgroup_kn_lock_live(of->kn, false);
if (!cgrp)
return -ENODEV;
spin_lock(&release_agent_path_lock);
- strlcpy(cgrp->root->release_agent_path, strstrip(buf),
+ strscpy(cgrp->root->release_agent_path, strstrip(buf),
sizeof(cgrp->root->release_agent_path));
spin_unlock(&release_agent_path_lock);
cgroup_kn_unlock(of->kn);
@@ -658,19 +677,19 @@ int proc_cgroupstats_show(struct seq_file *m, void *v)
seq_puts(m, "#subsys_name\thierarchy\tnum_cgroups\tenabled\n");
/*
- * ideally we don't want subsystems moving around while we do this.
- * cgroup_mutex is also necessary to guarantee an atomic snapshot of
- * subsys/hierarchy state.
+ * Grab the subsystems state racily. No need to add avenue to
+ * cgroup_mutex contention.
*/
- mutex_lock(&cgroup_mutex);
- for_each_subsys(ss, i)
+ for_each_subsys(ss, i) {
+ if (cgroup1_subsys_absent(ss))
+ continue;
seq_printf(m, "%s\t%d\t%d\t%d\n",
ss->legacy_name, ss->root->hierarchy_id,
atomic_read(&ss->root->nr_cgrps),
cgroup_ssid_enabled(i));
+ }
- mutex_unlock(&cgroup_mutex);
return 0;
}
@@ -682,6 +701,8 @@ int proc_cgroupstats_show(struct seq_file *m, void *v)
*
* Build and fill cgroupstats so that taskstats can export it to user
* space.
+ *
+ * Return: %0 on success or a negative errno code on failure
*/
int cgroupstats_build(struct cgroupstats *stats, struct dentry *dentry)
{
@@ -695,8 +716,6 @@ int cgroupstats_build(struct cgroupstats *stats, struct dentry *dentry)
kernfs_type(kn) != KERNFS_DIR)
return -EINVAL;
- mutex_lock(&cgroup_mutex);
-
/*
* We aren't being called from kernfs and there's no guarantee on
* @kn->priv's validity. For this and css_tryget_online_from_dir(),
@@ -704,16 +723,15 @@ int cgroupstats_build(struct cgroupstats *stats, struct dentry *dentry)
*/
rcu_read_lock();
cgrp = rcu_dereference(*(void __rcu __force **)&kn->priv);
- if (!cgrp || cgroup_is_dead(cgrp)) {
+ if (!cgrp || !cgroup_tryget(cgrp)) {
rcu_read_unlock();
- mutex_unlock(&cgroup_mutex);
return -ENOENT;
}
rcu_read_unlock();
css_task_iter_start(&cgrp->self, 0, &it);
while ((tsk = css_task_iter_next(&it))) {
- switch (tsk->state) {
+ switch (READ_ONCE(tsk->__state)) {
case TASK_RUNNING:
stats->nr_running++;
break;
@@ -727,14 +745,14 @@ int cgroupstats_build(struct cgroupstats *stats, struct dentry *dentry)
stats->nr_stopped++;
break;
default:
- if (delayacct_is_task_waiting_on_io(tsk))
+ if (tsk->in_iowait)
stats->nr_io_wait++;
break;
}
}
css_task_iter_end(&it);
- mutex_unlock(&cgroup_mutex);
+ cgroup_put(cgrp);
return 0;
}
@@ -787,13 +805,13 @@ void cgroup1_release_agent(struct work_struct *work)
goto out_free;
spin_lock(&release_agent_path_lock);
- strlcpy(agentbuf, cgrp->root->release_agent_path, PATH_MAX);
+ strscpy(agentbuf, cgrp->root->release_agent_path, PATH_MAX);
spin_unlock(&release_agent_path_lock);
if (!agentbuf[0])
goto out_free;
ret = cgroup_path_ns(cgrp, pathbuf, PATH_MAX, &init_cgroup_ns);
- if (ret < 0 || ret >= PATH_MAX)
+ if (ret < 0)
goto out_free;
argv[0] = agentbuf;
@@ -820,6 +838,10 @@ static int cgroup1_rename(struct kernfs_node *kn, struct kernfs_node *new_parent
struct cgroup *cgrp = kn->priv;
int ret;
+ /* do not accept '\n' to prevent making /proc/<pid>/cgroup unparsable */
+ if (strchr(new_name_str, '\n'))
+ return -EINVAL;
+
if (kernfs_type(kn) != KERNFS_DIR)
return -ENOTDIR;
if (kn->parent != new_parent)
@@ -833,13 +855,13 @@ static int cgroup1_rename(struct kernfs_node *kn, struct kernfs_node *new_parent
kernfs_break_active_protection(new_parent);
kernfs_break_active_protection(kn);
- mutex_lock(&cgroup_mutex);
+ cgroup_lock();
ret = kernfs_rename(kn, new_parent, new_name_str);
if (!ret)
TRACE_CGROUP_PATH(rename, cgrp);
- mutex_unlock(&cgroup_mutex);
+ cgroup_unlock();
kernfs_unbreak_active_protection(kn);
kernfs_unbreak_active_protection(new_parent);
@@ -861,6 +883,8 @@ static int cgroup1_show_options(struct seq_file *seq, struct kernfs_root *kf_roo
seq_puts(seq, ",xattr");
if (root->flags & CGRP_ROOT_CPUSET_V2_MODE)
seq_puts(seq, ",cpuset_v2_mode");
+ if (root->flags & CGRP_ROOT_FAVOR_DYNMODS)
+ seq_puts(seq, ",favordynmods");
spin_lock(&release_agent_path_lock);
if (strlen(root->release_agent_path))
@@ -884,6 +908,8 @@ enum cgroup1_param {
Opt_noprefix,
Opt_release_agent,
Opt_xattr,
+ Opt_favordynmods,
+ Opt_nofavordynmods,
};
const struct fs_parameter_spec cgroup1_fs_parameters[] = {
@@ -895,6 +921,8 @@ const struct fs_parameter_spec cgroup1_fs_parameters[] = {
fsparam_flag ("noprefix", Opt_noprefix),
fsparam_string("release_agent", Opt_release_agent),
fsparam_flag ("xattr", Opt_xattr),
+ fsparam_flag ("favordynmods", Opt_favordynmods),
+ fsparam_flag ("nofavordynmods", Opt_nofavordynmods),
{}
};
@@ -907,15 +935,14 @@ int cgroup1_parse_param(struct fs_context *fc, struct fs_parameter *param)
opt = fs_parse(fc, cgroup1_fs_parameters, param, &result);
if (opt == -ENOPARAM) {
- if (strcmp(param->key, "source") == 0) {
- if (fc->source)
- return invalf(fc, "Multiple sources not supported");
- fc->source = param->string;
- param->string = NULL;
- return 0;
- }
+ int ret;
+
+ ret = vfs_parse_fs_param_source(fc, param);
+ if (ret != -ENOPARAM)
+ return ret;
for_each_subsys(ss, i) {
- if (strcmp(param->key, ss->legacy_name))
+ if (strcmp(param->key, ss->legacy_name) ||
+ cgroup1_subsys_absent(ss))
continue;
if (!cgroup_ssid_enabled(i) || cgroup1_ssid_disabled(i))
return invalfc(fc, "Disabled controller '%s'",
@@ -948,10 +975,22 @@ int cgroup1_parse_param(struct fs_context *fc, struct fs_parameter *param)
case Opt_xattr:
ctx->flags |= CGRP_ROOT_XATTR;
break;
+ case Opt_favordynmods:
+ ctx->flags |= CGRP_ROOT_FAVOR_DYNMODS;
+ break;
+ case Opt_nofavordynmods:
+ ctx->flags &= ~CGRP_ROOT_FAVOR_DYNMODS;
+ break;
case Opt_release_agent:
/* Specifying two release agents is forbidden */
if (ctx->release_agent)
return invalfc(fc, "release_agent respecified");
+ /*
+ * Release agent gets called with all capabilities,
+ * require capabilities to set release agent.
+ */
+ if ((fc->user_ns != &init_user_ns) || !capable(CAP_SYS_ADMIN))
+ return invalfc(fc, "Setting release_agent not allowed");
ctx->release_agent = param->string;
param->string = NULL;
break;
@@ -995,13 +1034,14 @@ static int check_cgroupfs_options(struct fs_context *fc)
mask = ~((u16)1 << cpuset_cgrp_id);
#endif
for_each_subsys(ss, i)
- if (cgroup_ssid_enabled(i) && !cgroup1_ssid_disabled(i))
+ if (cgroup_ssid_enabled(i) && !cgroup1_ssid_disabled(i) &&
+ !cgroup1_subsys_absent(ss))
enabled |= 1 << i;
ctx->subsys_mask &= enabled;
/*
- * In absense of 'none', 'name=' or subsystem name options,
+ * In absence of 'none', 'name=' and subsystem name options,
* let's default to 'all'.
*/
if (!ctx->subsys_mask && !ctx->none && !ctx->name)
@@ -1089,7 +1129,7 @@ int cgroup1_reconfigure(struct fs_context *fc)
trace_cgroup_remount(root);
out_unlock:
- mutex_unlock(&cgroup_mutex);
+ cgroup_unlock();
return ret;
}
@@ -1193,8 +1233,11 @@ static int cgroup1_root_to_use(struct fs_context *fc)
init_cgroup_root(ctx);
ret = cgroup_setup_root(root, ctx->subsys_mask);
- if (ret)
+ if (!ret)
+ cgroup_favor_dynmods(root, ctx->flags & CGRP_ROOT_FAVOR_DYNMODS);
+ else
cgroup_free_root(root);
+
return ret;
}
@@ -1213,15 +1256,13 @@ int cgroup1_get_tree(struct fs_context *fc)
if (!ret && !percpu_ref_tryget_live(&ctx->root->cgrp.self.refcnt))
ret = 1; /* restart */
- mutex_unlock(&cgroup_mutex);
+ cgroup_unlock();
if (!ret)
ret = cgroup_do_get_tree(fc);
if (!ret && percpu_ref_is_dying(&ctx->root->cgrp.self.refcnt)) {
- struct super_block *sb = fc->root->d_sb;
- dput(fc->root);
- deactivate_locked_super(sb);
+ fc_drop_locked(fc);
ret = 1;
}
@@ -1232,6 +1273,40 @@ int cgroup1_get_tree(struct fs_context *fc)
return ret;
}
+/**
+ * task_get_cgroup1 - Acquires the associated cgroup of a task within a
+ * specific cgroup1 hierarchy. The cgroup1 hierarchy is identified by its
+ * hierarchy ID.
+ * @tsk: The target task
+ * @hierarchy_id: The ID of a cgroup1 hierarchy
+ *
+ * On success, the cgroup is returned. On failure, ERR_PTR is returned.
+ * We limit it to cgroup1 only.
+ */
+struct cgroup *task_get_cgroup1(struct task_struct *tsk, int hierarchy_id)
+{
+ struct cgroup *cgrp = ERR_PTR(-ENOENT);
+ struct cgroup_root *root;
+ unsigned long flags;
+
+ rcu_read_lock();
+ for_each_root(root) {
+ /* cgroup1 only*/
+ if (root == &cgrp_dfl_root)
+ continue;
+ if (root->hierarchy_id != hierarchy_id)
+ continue;
+ spin_lock_irqsave(&css_set_lock, flags);
+ cgrp = task_cgroup_from_root(tsk, root);
+ if (!cgrp || !cgroup_tryget(cgrp))
+ cgrp = ERR_PTR(-ENOENT);
+ spin_unlock_irqrestore(&css_set_lock, flags);
+ break;
+ }
+ rcu_read_unlock();
+ return cgrp;
+}
+
static int __init cgroup1_wq_init(void)
{
/*
@@ -1271,6 +1346,7 @@ static int __init cgroup_no_v1(char *str)
continue;
cgroup_no_v1_mask |= 1 << i;
+ break;
}
}
return 1;
generated by cgit (git 2.34.1) at 2025-07-24 08:05:59 +0000