summaryrefslogtreecommitdiffstats
path: root/kernel/cgroup
diff options
context:
space:
mode:
Diffstat (limited to 'kernel/cgroup')
-rw-r--r--kernel/cgroup/cgroup-internal.h19
-rw-r--r--kernel/cgroup/cgroup-v1.c33
-rw-r--r--kernel/cgroup/cgroup.c95
-rw-r--r--kernel/cgroup/cpuset.c16
-rw-r--r--kernel/cgroup/rstat.c51
5 files changed, 129 insertions, 85 deletions
diff --git a/kernel/cgroup/cgroup-internal.h b/kernel/cgroup/cgroup-internal.h
index bfbeabc17a9d..6e36e854b512 100644
--- a/kernel/cgroup/cgroup-internal.h
+++ b/kernel/cgroup/cgroup-internal.h
@@ -65,6 +65,25 @@ static inline struct cgroup_fs_context *cgroup_fc2context(struct fs_context *fc)
return container_of(kfc, struct cgroup_fs_context, kfc);
}
+struct cgroup_pidlist;
+
+struct cgroup_file_ctx {
+ struct cgroup_namespace *ns;
+
+ struct {
+ void *trigger;
+ } psi;
+
+ struct {
+ bool started;
+ struct css_task_iter iter;
+ } procs;
+
+ struct {
+ struct cgroup_pidlist *pidlist;
+ } procs1;
+};
+
/*
* A cgroup can be associated with multiple css_sets as different tasks may
* belong to different cgroups on different hierarchies. In the other
diff --git a/kernel/cgroup/cgroup-v1.c b/kernel/cgroup/cgroup-v1.c
index 81c9e0685948..41e0837a5a0b 100644
--- a/kernel/cgroup/cgroup-v1.c
+++ b/kernel/cgroup/cgroup-v1.c
@@ -394,6 +394,7 @@ 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;
@@ -403,25 +404,24 @@ 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;
@@ -449,7 +449,8 @@ 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,
@@ -460,7 +461,8 @@ 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;
/*
@@ -504,10 +506,11 @@ 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) &&
diff --git a/kernel/cgroup/cgroup.c b/kernel/cgroup/cgroup.c
index 919194de39c8..b31e1465868a 100644
--- a/kernel/cgroup/cgroup.c
+++ b/kernel/cgroup/cgroup.c
@@ -30,6 +30,7 @@
#include "cgroup-internal.h"
+#include <linux/bpf-cgroup.h>
#include <linux/cred.h>
#include <linux/errno.h>
#include <linux/init_task.h>
@@ -2650,11 +2651,11 @@ void cgroup_migrate_add_src(struct css_set *src_cset,
if (src_cset->dead)
return;
- src_cgrp = cset_cgroup_from_root(src_cset, dst_cgrp->root);
-
if (!list_empty(&src_cset->mg_preload_node))
return;
+ src_cgrp = cset_cgroup_from_root(src_cset, dst_cgrp->root);
+
WARN_ON(src_cset->mg_src_cgrp);
WARN_ON(src_cset->mg_dst_cgrp);
WARN_ON(!list_empty(&src_cset->mg_tasks));
@@ -3630,6 +3631,7 @@ static int cgroup_cpu_pressure_show(struct seq_file *seq, void *v)
static ssize_t cgroup_pressure_write(struct kernfs_open_file *of, char *buf,
size_t nbytes, enum psi_res res)
{
+ struct cgroup_file_ctx *ctx = of->priv;
struct psi_trigger *new;
struct cgroup *cgrp;
struct psi_group *psi;
@@ -3648,7 +3650,7 @@ static ssize_t cgroup_pressure_write(struct kernfs_open_file *of, char *buf,
return PTR_ERR(new);
}
- psi_trigger_replace(&of->priv, new);
+ psi_trigger_replace(&ctx->psi.trigger, new);
cgroup_put(cgrp);
@@ -3679,12 +3681,16 @@ static ssize_t cgroup_cpu_pressure_write(struct kernfs_open_file *of,
static __poll_t cgroup_pressure_poll(struct kernfs_open_file *of,
poll_table *pt)
{
- return psi_trigger_poll(&of->priv, of->file, pt);
+ struct cgroup_file_ctx *ctx = of->priv;
+
+ return psi_trigger_poll(&ctx->psi.trigger, of->file, pt);
}
static void cgroup_pressure_release(struct kernfs_open_file *of)
{
- psi_trigger_replace(&of->priv, NULL);
+ struct cgroup_file_ctx *ctx = of->priv;
+
+ psi_trigger_replace(&ctx->psi.trigger, NULL);
}
bool cgroup_psi_enabled(void)
@@ -3811,24 +3817,43 @@ static ssize_t cgroup_kill_write(struct kernfs_open_file *of, char *buf,
static int cgroup_file_open(struct kernfs_open_file *of)
{
struct cftype *cft = of_cft(of);
+ struct cgroup_file_ctx *ctx;
+ int ret;
- if (cft->open)
- return cft->open(of);
- return 0;
+ ctx = kzalloc(sizeof(*ctx), GFP_KERNEL);
+ if (!ctx)
+ return -ENOMEM;
+
+ ctx->ns = current->nsproxy->cgroup_ns;
+ get_cgroup_ns(ctx->ns);
+ of->priv = ctx;
+
+ if (!cft->open)
+ return 0;
+
+ ret = cft->open(of);
+ if (ret) {
+ put_cgroup_ns(ctx->ns);
+ kfree(ctx);
+ }
+ return ret;
}
static void cgroup_file_release(struct kernfs_open_file *of)
{
struct cftype *cft = of_cft(of);
+ struct cgroup_file_ctx *ctx = of->priv;
if (cft->release)
cft->release(of);
+ put_cgroup_ns(ctx->ns);
+ kfree(ctx);
}
static ssize_t cgroup_file_write(struct kernfs_open_file *of, char *buf,
size_t nbytes, loff_t off)
{
- struct cgroup_namespace *ns = current->nsproxy->cgroup_ns;
+ struct cgroup_file_ctx *ctx = of->priv;
struct cgroup *cgrp = of->kn->parent->priv;
struct cftype *cft = of_cft(of);
struct cgroup_subsys_state *css;
@@ -3845,7 +3870,7 @@ static ssize_t cgroup_file_write(struct kernfs_open_file *of, char *buf,
*/
if ((cgrp->root->flags & CGRP_ROOT_NS_DELEGATE) &&
!(cft->flags & CFTYPE_NS_DELEGATABLE) &&
- ns != &init_cgroup_ns && ns->root_cset->dfl_cgrp == cgrp)
+ ctx->ns != &init_cgroup_ns && ctx->ns->root_cset->dfl_cgrp == cgrp)
return -EPERM;
if (cft->write)
@@ -4751,21 +4776,21 @@ void css_task_iter_end(struct css_task_iter *it)
static void cgroup_procs_release(struct kernfs_open_file *of)
{
- if (of->priv) {
- css_task_iter_end(of->priv);
- kfree(of->priv);
- }
+ struct cgroup_file_ctx *ctx = of->priv;
+
+ if (ctx->procs.started)
+ css_task_iter_end(&ctx->procs.iter);
}
static void *cgroup_procs_next(struct seq_file *s, void *v, loff_t *pos)
{
struct kernfs_open_file *of = s->private;
- struct css_task_iter *it = of->priv;
+ struct cgroup_file_ctx *ctx = of->priv;
if (pos)
(*pos)++;
- return css_task_iter_next(it);
+ return css_task_iter_next(&ctx->procs.iter);
}
static void *__cgroup_procs_start(struct seq_file *s, loff_t *pos,
@@ -4773,21 +4798,18 @@ static void *__cgroup_procs_start(struct seq_file *s, loff_t *pos,
{
struct kernfs_open_file *of = s->private;
struct cgroup *cgrp = seq_css(s)->cgroup;
- struct css_task_iter *it = of->priv;
+ struct cgroup_file_ctx *ctx = of->priv;
+ struct css_task_iter *it = &ctx->procs.iter;
/*
* When a seq_file is seeked, it's always traversed sequentially
* from position 0, so we can simply keep iterating on !0 *pos.
*/
- if (!it) {
+ if (!ctx->procs.started) {
if (WARN_ON_ONCE((*pos)))
return ERR_PTR(-EINVAL);
-
- it = kzalloc(sizeof(*it), GFP_KERNEL);
- if (!it)
- return ERR_PTR(-ENOMEM);
- of->priv = it;
css_task_iter_start(&cgrp->self, iter_flags, it);
+ ctx->procs.started = true;
} else if (!(*pos)) {
css_task_iter_end(it);
css_task_iter_start(&cgrp->self, iter_flags, it);
@@ -4838,9 +4860,9 @@ static int cgroup_may_write(const struct cgroup *cgrp, struct super_block *sb)
static int cgroup_procs_write_permission(struct cgroup *src_cgrp,
struct cgroup *dst_cgrp,
- struct super_block *sb)
+ struct super_block *sb,
+ struct cgroup_namespace *ns)
{
- struct cgroup_namespace *ns = current->nsproxy->cgroup_ns;
struct cgroup *com_cgrp = src_cgrp;
int ret;
@@ -4869,11 +4891,12 @@ static int cgroup_procs_write_permission(struct cgroup *src_cgrp,
static int cgroup_attach_permissions(struct cgroup *src_cgrp,
struct cgroup *dst_cgrp,
- struct super_block *sb, bool threadgroup)
+ struct super_block *sb, bool threadgroup,
+ struct cgroup_namespace *ns)
{
int ret = 0;
- ret = cgroup_procs_write_permission(src_cgrp, dst_cgrp, sb);
+ ret = cgroup_procs_write_permission(src_cgrp, dst_cgrp, sb, ns);
if (ret)
return ret;
@@ -4890,8 +4913,10 @@ static int cgroup_attach_permissions(struct cgroup *src_cgrp,
static ssize_t __cgroup_procs_write(struct kernfs_open_file *of, char *buf,
bool threadgroup)
{
+ struct cgroup_file_ctx *ctx = of->priv;
struct cgroup *src_cgrp, *dst_cgrp;
struct task_struct *task;
+ const struct cred *saved_cred;
ssize_t ret;
bool locked;
@@ -4909,9 +4934,16 @@ static ssize_t __cgroup_procs_write(struct kernfs_open_file *of, char *buf,
src_cgrp = task_cgroup_from_root(task, &cgrp_dfl_root);
spin_unlock_irq(&css_set_lock);
- /* process and thread migrations follow same delegation rule */
+ /*
+ * Process and thread migrations follow same delegation rule. Check
+ * permissions using the credentials from file open to protect against
+ * inherited fd attacks.
+ */
+ saved_cred = override_creds(of->file->f_cred);
ret = cgroup_attach_permissions(src_cgrp, dst_cgrp,
- of->file->f_path.dentry->d_sb, threadgroup);
+ of->file->f_path.dentry->d_sb,
+ threadgroup, ctx->ns);
+ revert_creds(saved_cred);
if (ret)
goto out_finish;
@@ -5711,7 +5743,7 @@ static void __init cgroup_init_subsys(struct cgroup_subsys *ss, bool early)
/* Create the root cgroup state for this subsystem */
ss->root = &cgrp_dfl_root;
- css = ss->css_alloc(cgroup_css(&cgrp_dfl_root.cgrp, ss));
+ css = ss->css_alloc(NULL);
/* We don't handle early failures gracefully */
BUG_ON(IS_ERR(css));
init_and_link_css(css, ss, &cgrp_dfl_root.cgrp);
@@ -6130,7 +6162,8 @@ static int cgroup_css_set_fork(struct kernel_clone_args *kargs)
goto err;
ret = cgroup_attach_permissions(cset->dfl_cgrp, dst_cgrp, sb,
- !(kargs->flags & CLONE_THREAD));
+ !(kargs->flags & CLONE_THREAD),
+ current->nsproxy->cgroup_ns);
if (ret)
goto err;
diff --git a/kernel/cgroup/cpuset.c b/kernel/cgroup/cpuset.c
index d0e163a02099..dc653ab26e50 100644
--- a/kernel/cgroup/cpuset.c
+++ b/kernel/cgroup/cpuset.c
@@ -616,19 +616,11 @@ static int validate_change(struct cpuset *cur, struct cpuset *trial)
struct cpuset *c, *par;
int ret;
- rcu_read_lock();
-
- /* Each of our child cpusets must be a subset of us */
- ret = -EBUSY;
- cpuset_for_each_child(c, css, cur)
- if (!is_cpuset_subset(c, trial))
- goto out;
-
- /* Remaining checks don't apply to root cpuset */
- ret = 0;
+ /* The checks don't apply to root cpuset */
if (cur == &top_cpuset)
- goto out;
+ return 0;
+ rcu_read_lock();
par = parent_cs(cur);
/* On legacy hierarchy, we must be a subset of our parent cpuset. */
@@ -3536,7 +3528,7 @@ static struct cpuset *nearest_hardwall_ancestor(struct cpuset *cs)
bool __cpuset_node_allowed(int node, gfp_t gfp_mask)
{
struct cpuset *cs; /* current cpuset ancestors */
- int allowed; /* is allocation in zone z allowed? */
+ bool allowed; /* is allocation in zone z allowed? */
unsigned long flags;
if (in_interrupt())
diff --git a/kernel/cgroup/rstat.c b/kernel/cgroup/rstat.c
index 1486768f2318..9d331ba44870 100644
--- a/kernel/cgroup/rstat.c
+++ b/kernel/cgroup/rstat.c
@@ -35,7 +35,7 @@ void cgroup_rstat_updated(struct cgroup *cgrp, int cpu)
* instead of NULL, we can tell whether @cgrp is on the list by
* testing the next pointer for NULL.
*/
- if (cgroup_rstat_cpu(cgrp, cpu)->updated_next)
+ if (data_race(cgroup_rstat_cpu(cgrp, cpu)->updated_next))
return;
raw_spin_lock_irqsave(cpu_lock, flags);
@@ -88,6 +88,7 @@ static struct cgroup *cgroup_rstat_cpu_pop_updated(struct cgroup *pos,
struct cgroup *root, int cpu)
{
struct cgroup_rstat_cpu *rstatc;
+ struct cgroup *parent;
if (pos == root)
return NULL;
@@ -96,10 +97,14 @@ static struct cgroup *cgroup_rstat_cpu_pop_updated(struct cgroup *pos,
* We're gonna walk down to the first leaf and visit/remove it. We
* can pick whatever unvisited node as the starting point.
*/
- if (!pos)
+ if (!pos) {
pos = root;
- else
+ /* return NULL if this subtree is not on-list */
+ if (!cgroup_rstat_cpu(pos, cpu)->updated_next)
+ return NULL;
+ } else {
pos = cgroup_parent(pos);
+ }
/* walk down to the first leaf */
while (true) {
@@ -115,33 +120,25 @@ static struct cgroup *cgroup_rstat_cpu_pop_updated(struct cgroup *pos,
* However, due to the way we traverse, @pos will be the first
* child in most cases. The only exception is @root.
*/
- if (rstatc->updated_next) {
- struct cgroup *parent = cgroup_parent(pos);
-
- if (parent) {
- struct cgroup_rstat_cpu *prstatc;
- struct cgroup **nextp;
-
- prstatc = cgroup_rstat_cpu(parent, cpu);
- nextp = &prstatc->updated_children;
- while (true) {
- struct cgroup_rstat_cpu *nrstatc;
-
- nrstatc = cgroup_rstat_cpu(*nextp, cpu);
- if (*nextp == pos)
- break;
- WARN_ON_ONCE(*nextp == parent);
- nextp = &nrstatc->updated_next;
- }
- *nextp = rstatc->updated_next;
- }
+ parent = cgroup_parent(pos);
+ if (parent) {
+ struct cgroup_rstat_cpu *prstatc;
+ struct cgroup **nextp;
- rstatc->updated_next = NULL;
- return pos;
+ prstatc = cgroup_rstat_cpu(parent, cpu);
+ nextp = &prstatc->updated_children;
+ while (*nextp != pos) {
+ struct cgroup_rstat_cpu *nrstatc;
+
+ nrstatc = cgroup_rstat_cpu(*nextp, cpu);
+ WARN_ON_ONCE(*nextp == parent);
+ nextp = &nrstatc->updated_next;
+ }
+ *nextp = rstatc->updated_next;
}
- /* only happens for @root */
- return NULL;
+ rstatc->updated_next = NULL;
+ return pos;
}
/* see cgroup_rstat_flush() */