From: Waiman Long <longman@redhat.com>
To: "Ridong Chen" <ridong.chen@linux.dev>,
"Tejun Heo" <tj@kernel.org>,
"Johannes Weiner" <hannes@cmpxchg.org>,
"Michal Koutný" <mkoutny@suse.com>,
"Li Zefan" <lizefan@huawei.com>,
"Farhad Alemi" <farhad.alemi@berkeley.edu>,
"Andrew Morton" <akpm@linux-foundation.org>
Cc: cgroups@vger.kernel.org, linux-kernel@vger.kernel.org,
Aaron Tomlin <atomlin@atomlin.com>,
Guopeng Zhang <guopeng.zhang@linux.dev>,
Gregory Price <gourry@gourry.net>,
David Hildenbrand <david@kernel.org>,
Waiman Long <longman@redhat.com>
Subject: [PATCH v7 8/9] cgroup/cpuset: Support multiple source cpusets for cpuset_*attach()
Date: Sat, 20 Jun 2026 23:28:15 -0400 [thread overview]
Message-ID: <20260621032816.1806773-9-longman@redhat.com> (raw)
In-Reply-To: <20260621032816.1806773-1-longman@redhat.com>
There are 2 possible scenarios where the cgroup_taskset structure
passed into the cgroup can_attach() and attach() methods can contain
task migration data with multiple source cpusets.
- A multithread application with threads in different cpusets is
fully migrated into a new cpuset.
- Disabling v2 cpuset controller will move all the tasks in child
cpusets to the parent cpuset.
The current cpuset_can_attach() and cpuset_attach() functions still
expect task migration is from one source cpuset to one destination
cpuset.
Fix that by tracking the set of source (old) cpusets in singly linked
lists with the setting of attach_in_progress flag associated with the
insertion into the list. The list will be iterated when necessary to
properly update the internal data.
To ensure proper DL tasks accounting, the nr_migrate_dl_tasks in both
the source and destination cpusets are decremented/incremented with
their values added to nr_deadline_tasks when the migration is successful.
The setting of the global attach_cpus_updated and attach_mems_updated
flags are also moved from cpuset_attach() to cpuset_can_attach() as the
correct source cpuset can no longer be determined in cpuset_attach()
and cpuset states will not be changed between cpuset_attach() and
cpuset_can_attach() with an earlier patch.
Signed-off-by: Waiman Long <longman@redhat.com>
---
kernel/cgroup/cpuset-internal.h | 1 +
kernel/cgroup/cpuset.c | 66 ++++++++++++++++++++++++++++-----
2 files changed, 57 insertions(+), 10 deletions(-)
diff --git a/kernel/cgroup/cpuset-internal.h b/kernel/cgroup/cpuset-internal.h
index f7aaf01f7cd5..011993b1f756 100644
--- a/kernel/cgroup/cpuset-internal.h
+++ b/kernel/cgroup/cpuset-internal.h
@@ -149,6 +149,7 @@ struct cpuset {
* Tasks are being attached to this cpuset. Used to prevent
* zeroing cpus/mems_allowed between ->can_attach() and ->attach().
*/
+ struct llist_node attach_node;
int attach_in_progress;
/* partition root state */
diff --git a/kernel/cgroup/cpuset.c b/kernel/cgroup/cpuset.c
index 511afb077e2d..c2d172873166 100644
--- a/kernel/cgroup/cpuset.c
+++ b/kernel/cgroup/cpuset.c
@@ -37,6 +37,7 @@
#include <linux/wait.h>
#include <linux/workqueue.h>
#include <linux/task_work.h>
+#include <linux/llist.h>
DEFINE_STATIC_KEY_FALSE(cpusets_pre_enable_key);
DEFINE_STATIC_KEY_FALSE(cpusets_enabled_key);
@@ -584,6 +585,7 @@ static struct cpuset *dup_or_alloc_cpuset(struct cpuset *cs)
return NULL;
trial->dl_bw_cpu = -1;
+ init_llist_node(&trial->attach_node);
/* Setup cpumask pointer array */
cpumask_var_t *pmask[4] = {
@@ -2983,9 +2985,10 @@ static int update_prstate(struct cpuset *cs, int new_prs)
* Protected by cpuset_mutex
*
* The attach_cpus_updated/attach_mems_updated flags are set in either
- * cpuset_attach() or cpuset_fork() and used in cpuset_attach_task().
+ * cpuset_can_attach() or cpuset_fork() and used in cpuset_attach_task().
*/
static struct cpuset *cpuset_attach_old_cs;
+static LLIST_HEAD(src_cs_head);
static bool attach_cpus_updated;
static bool attach_mems_updated;
@@ -3001,6 +3004,8 @@ static bool attach_mems_updated;
static int cpuset_can_attach_check(struct cpuset *cs, struct cpuset *oldcs,
bool *psetsched)
{
+ bool cpus_updated, mems_updated;
+
if (cpumask_empty(cs->effective_cpus) ||
(!is_in_v2_mode() && nodes_empty(cs->mems_allowed)))
return -ENOSPC;
@@ -3008,14 +3013,25 @@ static int cpuset_can_attach_check(struct cpuset *cs, struct cpuset *oldcs,
if (!oldcs)
return 0;
+ if (!llist_on_list(&oldcs->attach_node)) {
+ llist_add(&oldcs->attach_node, &src_cs_head);
+ oldcs->attach_in_progress++;
+ }
+
+ cpus_updated = !cpumask_equal(cs->effective_cpus, oldcs->effective_cpus);
+ mems_updated = !nodes_equal(cs->effective_mems, oldcs->effective_mems);
+
+ if (cpus_updated)
+ attach_cpus_updated = true;
+ if (mems_updated)
+ attach_mems_updated = true;
+
/*
* Skip rights over task setsched check in v2 when nothing changes,
* migration permission derives from hierarchy ownership in
* cgroup_procs_write_permission()).
*/
- *psetsched = !cpuset_v2() ||
- !cpumask_equal(cs->effective_cpus, oldcs->effective_cpus) ||
- !nodes_equal(cs->effective_mems, oldcs->effective_mems);
+ *psetsched = !cpuset_v2() || cpus_updated || mems_updated;
/*
* A v1 cpuset with tasks will have no CPU left only when CPU hotplug
@@ -3056,6 +3072,26 @@ static void reset_migrate_dl_data(struct cpuset *cs)
cs->dl_bw_cpu = -1;
}
+/*
+ * Clear and optionally apply (@cancel is false) the attach related data in the
+ * source cpusets.
+ */
+static void clear_attach_data(struct llist_head *head, bool cancel)
+{
+ struct cpuset *cs, *next;
+ struct llist_node *lnode = __llist_del_all(head);
+
+ llist_for_each_entry_safe(cs, next, lnode, attach_node) {
+ init_llist_node(&cs->attach_node);
+ if (cs->nr_migrate_dl_tasks) {
+ if (!cancel)
+ cs->nr_deadline_tasks += cs->nr_migrate_dl_tasks;
+ cs->nr_migrate_dl_tasks = 0;
+ }
+ dec_attach_in_progress_locked(cs);
+ }
+}
+
/* Called by cgroups to determine if a cpuset is usable; cpuset_mutex held */
static int cpuset_can_attach(struct cgroup_taskset *tset)
{
@@ -3071,6 +3107,8 @@ static int cpuset_can_attach(struct cgroup_taskset *tset)
cs = css_cs(css);
mutex_lock(&cpuset_mutex);
+ attach_cpus_updated = false;
+ attach_mems_updated = false;
/* Check to see if task is allowed in the cpuset */
ret = cpuset_can_attach_check(cs, oldcs, &setsched_check);
@@ -3095,6 +3133,15 @@ static int cpuset_can_attach(struct cgroup_taskset *tset)
* selected as cpuset_attach_old_cs.
*/
cgroup_taskset_for_each(task, css, tset) {
+ struct cpuset *new_oldcs = task_cs(task);
+
+ if (new_oldcs != oldcs) {
+ oldcs = new_oldcs;
+ ret = cpuset_can_attach_check(cs, oldcs, &setsched_check);
+ if (ret)
+ goto out_unlock;
+ }
+
ret = task_can_attach(task);
if (ret)
goto out_unlock;
@@ -3116,6 +3163,7 @@ static int cpuset_can_attach(struct cgroup_taskset *tset)
* contribute to sum_migrate_dl_bw.
*/
cs->nr_migrate_dl_tasks++;
+ oldcs->nr_migrate_dl_tasks--;
if (dl_task_needs_bw_move(task, cs->effective_cpus))
cs->sum_migrate_dl_bw += task->dl.dl_bw;
}
@@ -3126,9 +3174,9 @@ static int cpuset_can_attach(struct cgroup_taskset *tset)
out_unlock:
if (ret) {
reset_migrate_dl_data(cs);
+ clear_attach_data(&src_cs_head, true);
} else {
cs->attach_in_progress++;
- oldcs->attach_in_progress++;
}
mutex_unlock(&cpuset_mutex);
@@ -3145,6 +3193,7 @@ static void cpuset_cancel_attach(struct cgroup_taskset *tset)
mutex_lock(&cpuset_mutex);
dec_attach_in_progress_locked(cs);
+ clear_attach_data(&src_cs_head, true);
if (cs->dl_bw_cpu >= 0)
dl_bw_free(cs->dl_bw_cpu, cs->sum_migrate_dl_bw);
@@ -3224,7 +3273,6 @@ static void cpuset_attach(struct cgroup_taskset *tset)
struct task_struct *task;
struct cgroup_subsys_state *css;
struct cpuset *cs;
- struct cpuset *oldcs = cpuset_attach_old_cs;
cgroup_taskset_first(tset, &css);
cs = css_cs(css);
@@ -3232,9 +3280,6 @@ static void cpuset_attach(struct cgroup_taskset *tset)
lockdep_assert_cpus_held(); /* see cgroup_attach_lock() */
mutex_lock(&cpuset_mutex);
queue_task_work = false;
-
- attach_cpus_updated = !cpumask_equal(cs->effective_cpus, oldcs->effective_cpus);
- attach_mems_updated = !nodes_equal(cs->effective_mems, oldcs->effective_mems);
guarantee_online_mems(cs, &cpuset_attach_nodemask_to);
/*
@@ -3256,10 +3301,10 @@ static void cpuset_attach(struct cgroup_taskset *tset)
if (cs->nr_migrate_dl_tasks) {
cs->nr_deadline_tasks += cs->nr_migrate_dl_tasks;
- oldcs->nr_deadline_tasks -= cs->nr_migrate_dl_tasks;
reset_migrate_dl_data(cs);
}
+ clear_attach_data(&src_cs_head, false);
dec_attach_in_progress_locked(cs);
mutex_unlock(&cpuset_mutex);
@@ -3777,6 +3822,7 @@ int __init cpuset_init(void)
cpumask_setall(top_cpuset.effective_xcpus);
cpumask_setall(top_cpuset.exclusive_cpus);
nodes_setall(top_cpuset.effective_mems);
+ init_llist_node(&top_cpuset.attach_node);
cpuset1_init(&top_cpuset);
--
2.54.0
next prev parent reply other threads:[~2026-06-21 3:29 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-06-21 3:28 [PATCH v7 0/9] cgroup/cpuset: Support multiple source/destination cpusets for cpuset_*attach() Waiman Long
2026-06-21 3:28 ` [PATCH v7 1/9] cgroup/cpuset: rebind mm mempolicy to effective_mems, not mems_allowed Waiman Long
2026-06-21 3:28 ` [PATCH v7 2/9] cgroup/cpuset: Fix node inconsistencies between cpuset_update_tasks_nodemask() and cpuset_attach() Waiman Long
2026-06-21 3:28 ` [PATCH v7 3/9] cgroup/cpuset: Prevent race between task attach and cpuset state change Waiman Long
2026-06-21 3:28 ` [PATCH v7 4/9] cgroup/cpuset: Add a cpuset_reserve_dl_bw() helper Waiman Long
2026-06-21 3:28 ` [PATCH v7 5/9] cgroup/cpuset: Expand the scope of cpuset_can_attach_check() Waiman Long
2026-06-21 3:28 ` [PATCH v7 6/9] cgroup/cpuset: Make cpuset_attach_old_cs track task group leaders Waiman Long
2026-06-21 3:28 ` [PATCH v7 7/9] cgroup/cpuset: Move mpol_rebind_mm/cpuset_migrate_mm() calls inside cpuset_attach_task() Waiman Long
2026-06-21 3:28 ` Waiman Long [this message]
2026-06-21 3:28 ` [PATCH v7 9/9] cgroup/cpuset: Support multiple destination cpusets for cpuset_*attach() Waiman Long
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20260621032816.1806773-9-longman@redhat.com \
--to=longman@redhat.com \
--cc=akpm@linux-foundation.org \
--cc=atomlin@atomlin.com \
--cc=cgroups@vger.kernel.org \
--cc=david@kernel.org \
--cc=farhad.alemi@berkeley.edu \
--cc=gourry@gourry.net \
--cc=guopeng.zhang@linux.dev \
--cc=hannes@cmpxchg.org \
--cc=linux-kernel@vger.kernel.org \
--cc=lizefan@huawei.com \
--cc=mkoutny@suse.com \
--cc=ridong.chen@linux.dev \
--cc=tj@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox