From: Tejun Heo <tj@kernel.org>
To: linux-kernel@vger.kernel.org, sched-ext@lists.linux.dev
Cc: void@manifault.com, arighi@nvidia.com, changwoo@igalia.com,
emil@etsalapatis.com, Tejun Heo <tj@kernel.org>
Subject: [PATCH 06/34] sched_ext: Reorganize enable/disable path for multi-scheduler support
Date: Wed, 4 Mar 2026 12:00:51 -1000 [thread overview]
Message-ID: <20260304220119.4095551-7-tj@kernel.org> (raw)
In-Reply-To: <20260304220119.4095551-1-tj@kernel.org>
In preparation for multiple scheduler support, reorganize the enable and
disable paths to make scheduler instances explicit. Extract
scx_root_disable() from scx_disable_workfn(). Rename scx_enable_workfn()
to scx_root_enable_workfn(). Change scx_disable() to take @sch parameter
and only queue disable_work if scx_claim_exit() succeeds for consistency.
Move exit_kind validation into scx_claim_exit(). The sysrq handler now
prints a message when no scheduler is loaded.
These changes don't materially affect user-visible behavior.
v2: Keep scx_enable() name as-is and only rename the workfn to
scx_root_enable_workfn(). Change scx_enable() return type to s32.
Signed-off-by: Tejun Heo <tj@kernel.org>
---
kernel/sched/ext.c | 78 +++++++++++++++++++++++++---------------------
1 file changed, 43 insertions(+), 35 deletions(-)
diff --git a/kernel/sched/ext.c b/kernel/sched/ext.c
index d1f7de05da04..f1d946749e54 100644
--- a/kernel/sched/ext.c
+++ b/kernel/sched/ext.c
@@ -3267,8 +3267,8 @@ void sched_ext_dead(struct task_struct *p)
raw_spin_unlock_irqrestore(&scx_tasks_lock, flags);
/*
- * @p is off scx_tasks and wholly ours. scx_enable()'s READY -> ENABLED
- * transitions can't race us. Disable ops for @p.
+ * @p is off scx_tasks and wholly ours. scx_root_enable()'s READY ->
+ * ENABLED transitions can't race us. Disable ops for @p.
*/
if (scx_get_task_state(p) != SCX_TASK_NONE) {
struct rq_flags rf;
@@ -4430,24 +4430,12 @@ static void free_kick_syncs(void)
}
}
-static void scx_disable_workfn(struct kthread_work *work)
+static void scx_root_disable(struct scx_sched *sch)
{
- struct scx_sched *sch = container_of(work, struct scx_sched, disable_work);
struct scx_exit_info *ei = sch->exit_info;
struct scx_task_iter sti;
struct task_struct *p;
- int kind, cpu;
-
- kind = atomic_read(&sch->exit_kind);
- while (true) {
- if (kind == SCX_EXIT_DONE) /* already disabled? */
- return;
- WARN_ON_ONCE(kind == SCX_EXIT_NONE);
- if (atomic_try_cmpxchg(&sch->exit_kind, &kind, SCX_EXIT_DONE))
- break;
- }
- ei->kind = kind;
- ei->reason = scx_exit_reason(ei->kind);
+ int cpu;
/* guarantee forward progress by bypassing scx_ops */
scx_bypass(true);
@@ -4591,6 +4579,9 @@ static bool scx_claim_exit(struct scx_sched *sch, enum scx_exit_kind kind)
lockdep_assert_preemption_disabled();
+ if (WARN_ON_ONCE(kind == SCX_EXIT_NONE || kind == SCX_EXIT_DONE))
+ kind = SCX_EXIT_ERROR;
+
if (!atomic_try_cmpxchg(&sch->exit_kind, &none, kind))
return false;
@@ -4603,21 +4594,31 @@ static bool scx_claim_exit(struct scx_sched *sch, enum scx_exit_kind kind)
return true;
}
-static void scx_disable(enum scx_exit_kind kind)
+static void scx_disable_workfn(struct kthread_work *work)
{
- struct scx_sched *sch;
+ struct scx_sched *sch = container_of(work, struct scx_sched, disable_work);
+ struct scx_exit_info *ei = sch->exit_info;
+ int kind;
- if (WARN_ON_ONCE(kind == SCX_EXIT_NONE || kind == SCX_EXIT_DONE))
- kind = SCX_EXIT_ERROR;
+ kind = atomic_read(&sch->exit_kind);
+ while (true) {
+ if (kind == SCX_EXIT_DONE) /* already disabled? */
+ return;
+ WARN_ON_ONCE(kind == SCX_EXIT_NONE);
+ if (atomic_try_cmpxchg(&sch->exit_kind, &kind, SCX_EXIT_DONE))
+ break;
+ }
+ ei->kind = kind;
+ ei->reason = scx_exit_reason(ei->kind);
- rcu_read_lock();
- sch = rcu_dereference(scx_root);
- if (sch) {
- guard(preempt)();
- scx_claim_exit(sch, kind);
+ scx_root_disable(sch);
+}
+
+static void scx_disable(struct scx_sched *sch, enum scx_exit_kind kind)
+{
+ guard(preempt)();
+ if (scx_claim_exit(sch, kind))
kthread_queue_work(sch->helper, &sch->disable_work);
- }
- rcu_read_unlock();
}
static void dump_newline(struct seq_buf *s)
@@ -5135,10 +5136,9 @@ struct scx_enable_cmd {
int ret;
};
-static void scx_enable_workfn(struct kthread_work *work)
+static void scx_root_enable_workfn(struct kthread_work *work)
{
- struct scx_enable_cmd *cmd =
- container_of(work, struct scx_enable_cmd, work);
+ struct scx_enable_cmd *cmd = container_of(work, struct scx_enable_cmd, work);
struct sched_ext_ops *ops = cmd->ops;
struct scx_sched *sch;
struct scx_task_iter sti;
@@ -5387,12 +5387,12 @@ static void scx_enable_workfn(struct kthread_work *work)
* Flush scx_disable_work to ensure that error is reported before init
* completion. sch's base reference will be put by bpf_scx_unreg().
*/
- scx_error(sch, "scx_enable() failed (%d)", ret);
+ scx_error(sch, "scx_root_enable() failed (%d)", ret);
kthread_flush_work(&sch->disable_work);
cmd->ret = 0;
}
-static int scx_enable(struct sched_ext_ops *ops, struct bpf_link *link)
+static s32 scx_enable(struct sched_ext_ops *ops, struct bpf_link *link)
{
static struct kthread_worker *helper;
static DEFINE_MUTEX(helper_mutex);
@@ -5418,7 +5418,7 @@ static int scx_enable(struct sched_ext_ops *ops, struct bpf_link *link)
mutex_unlock(&helper_mutex);
}
- kthread_init_work(&cmd.work, scx_enable_workfn);
+ kthread_init_work(&cmd.work, scx_root_enable_workfn);
cmd.ops = ops;
kthread_queue_work(READ_ONCE(helper), &cmd.work);
@@ -5561,7 +5561,7 @@ static void bpf_scx_unreg(void *kdata, struct bpf_link *link)
struct sched_ext_ops *ops = kdata;
struct scx_sched *sch = ops->priv;
- scx_disable(SCX_EXIT_UNREG);
+ scx_disable(sch, SCX_EXIT_UNREG);
kthread_flush_work(&sch->disable_work);
kobject_put(&sch->kobj);
}
@@ -5689,7 +5689,15 @@ static struct bpf_struct_ops bpf_sched_ext_ops = {
static void sysrq_handle_sched_ext_reset(u8 key)
{
- scx_disable(SCX_EXIT_SYSRQ);
+ struct scx_sched *sch;
+
+ rcu_read_lock();
+ sch = rcu_dereference(scx_root);
+ if (likely(sch))
+ scx_disable(sch, SCX_EXIT_SYSRQ);
+ else
+ pr_info("sched_ext: BPF schedulers not loaded\n");
+ rcu_read_unlock();
}
static const struct sysrq_key_op sysrq_sched_ext_reset_op = {
--
2.53.0
next prev parent reply other threads:[~2026-03-04 22:01 UTC|newest]
Thread overview: 50+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-03-04 22:00 [PATCHSET v3 sched_ext/for-7.1] sched_ext: Implement cgroup sub-scheduler support Tejun Heo
2026-03-04 22:00 ` [PATCH 01/34] sched_ext: Implement cgroup subtree iteration for scx_task_iter Tejun Heo
2026-03-04 22:00 ` [PATCH 02/34] sched_ext: Add @kargs to scx_fork() Tejun Heo
2026-03-04 22:00 ` [PATCH 03/34] sched/core: Swap the order between sched_post_fork() and cgroup_post_fork() Tejun Heo
2026-03-06 4:17 ` Tejun Heo
2026-03-06 8:44 ` Peter Zijlstra
2026-03-04 22:00 ` [PATCH 04/34] cgroup: Expose some cgroup helpers Tejun Heo
2026-03-06 4:18 ` Tejun Heo
2026-03-04 22:00 ` [PATCH 05/34] sched_ext: Update p->scx.disallow warning in scx_init_task() Tejun Heo
2026-03-04 22:00 ` Tejun Heo [this message]
2026-03-04 22:00 ` [PATCH 07/34] sched_ext: Introduce cgroup sub-sched support Tejun Heo
2026-03-04 22:00 ` [PATCH 08/34] sched_ext: Introduce scx_task_sched[_rcu]() Tejun Heo
2026-03-04 22:00 ` [PATCH 09/34] sched_ext: Introduce scx_prog_sched() Tejun Heo
2026-03-04 22:00 ` [PATCH 10/34] sched_ext: Enforce scheduling authority in dispatch and select_cpu operations Tejun Heo
2026-03-04 22:00 ` [PATCH 11/34] sched_ext: Enforce scheduler ownership when updating slice and dsq_vtime Tejun Heo
2026-03-04 22:00 ` [PATCH 12/34] sched_ext: scx_dsq_move() should validate the task belongs to the right scheduler Tejun Heo
2026-03-04 22:00 ` [PATCH 13/34] sched_ext: Refactor task init/exit helpers Tejun Heo
2026-03-04 22:00 ` [PATCH 14/34] sched_ext: Make scx_prio_less() handle multiple schedulers Tejun Heo
2026-03-04 22:01 ` [PATCH 15/34] sched_ext: Move default slice to per-scheduler field Tejun Heo
2026-03-04 22:01 ` [PATCH 16/34] sched_ext: Move aborting flag " Tejun Heo
2026-03-04 22:01 ` [PATCH 17/34] sched_ext: Move bypass_dsq into scx_sched_pcpu Tejun Heo
2026-03-04 22:01 ` [PATCH 18/34] sched_ext: Move bypass state into scx_sched Tejun Heo
2026-03-04 22:01 ` [PATCH 19/34] sched_ext: Prepare bypass mode for hierarchical operation Tejun Heo
2026-03-04 22:01 ` [PATCH 20/34] sched_ext: Factor out scx_dispatch_sched() Tejun Heo
2026-03-04 22:01 ` [PATCH 21/34] sched_ext: When calling ops.dispatch() @prev must be on the same scx_sched Tejun Heo
2026-03-04 22:01 ` [PATCH 22/34] sched_ext: Separate bypass dispatch enabling from bypass depth tracking Tejun Heo
2026-03-04 22:01 ` [PATCH 23/34] sched_ext: Implement hierarchical bypass mode Tejun Heo
2026-03-06 7:03 ` Andrea Righi
2026-03-06 7:23 ` Andrea Righi
2026-03-06 17:39 ` [PATCH v2 " Tejun Heo
2026-03-04 22:01 ` [PATCH 24/34] sched_ext: Dispatch from all scx_sched instances Tejun Heo
2026-03-04 22:01 ` [PATCH 25/34] sched_ext: Move scx_dsp_ctx and scx_dsp_max_batch into scx_sched Tejun Heo
2026-03-04 22:01 ` [PATCH 26/34] sched_ext: Make watchdog sub-sched aware Tejun Heo
2026-03-04 22:01 ` [PATCH 27/34] sched_ext: Convert scx_dump_state() spinlock to raw spinlock Tejun Heo
2026-03-04 22:01 ` [PATCH 28/34] sched_ext: Support dumping multiple schedulers and add scheduler identification Tejun Heo
2026-03-04 22:01 ` [PATCH 29/34] sched_ext: Implement cgroup sub-sched enabling and disabling Tejun Heo
2026-03-06 9:41 ` Cheng-Yang Chou
2026-03-06 17:39 ` [PATCH v2 " Tejun Heo
2026-03-04 22:01 ` [PATCH 30/34] sched_ext: Add scx_sched back pointer to scx_sched_pcpu Tejun Heo
2026-03-04 22:01 ` [PATCH 31/34] sched_ext: Make scx_bpf_reenqueue_local() sub-sched aware Tejun Heo
2026-03-04 22:01 ` [PATCH 32/34] sched_ext: Factor out scx_link_sched() and scx_unlink_sched() Tejun Heo
2026-03-04 22:01 ` [PATCH 33/34] sched_ext: Add rhashtable lookup for sub-schedulers Tejun Heo
2026-03-04 22:01 ` [PATCH 34/34] sched_ext: Add basic building blocks for nested sub-scheduler dispatching Tejun Heo
2026-03-06 4:09 ` [PATCHSET v3 sched_ext/for-7.1] sched_ext: Implement cgroup sub-scheduler support Tejun Heo
2026-03-06 4:17 ` Tejun Heo
2026-03-06 7:29 ` Andrea Righi
2026-03-06 18:14 ` Tejun Heo
-- strict thread matches above, loose matches on Subject: below --
2026-02-25 5:01 [PATCHSET v2 " Tejun Heo
2026-02-25 5:01 ` [PATCH 06/34] sched_ext: Reorganize enable/disable path for multi-scheduler support Tejun Heo
2026-02-25 5:00 [PATCHSET v2 sched_ext/for-7.1] sched_ext: Implement cgroup sub-scheduler support Tejun Heo
2026-02-25 5:00 ` [PATCH 06/34] sched_ext: Reorganize enable/disable path for multi-scheduler support Tejun Heo
2026-01-21 23:11 [PATCHSET v1 sched_ext/for-6.20] sched_ext: Implement cgroup sub-scheduler support Tejun Heo
2026-01-21 23:11 ` [PATCH 06/34] sched_ext: Reorganize enable/disable path for multi-scheduler support Tejun Heo
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=20260304220119.4095551-7-tj@kernel.org \
--to=tj@kernel.org \
--cc=arighi@nvidia.com \
--cc=changwoo@igalia.com \
--cc=emil@etsalapatis.com \
--cc=linux-kernel@vger.kernel.org \
--cc=sched-ext@lists.linux.dev \
--cc=void@manifault.com \
/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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.