[PATCH] sched/core: Call wq_worker_tick() for the execution context

[PATCH] sched/core: Call wq_worker_tick() for the execution context

[Hui Su]

wq_worker_tick() accounts CPU time and detects CPU-intensive work for
the kworker that is actually running. With proxy execution, rq->donor
is the scheduling context while rq->curr is the execution context.

Calling the hook with rq->donor can skip workqueue accounting when a
kworker is executing on behalf of a donor task. It can also account a
blocked kworker when the donor is a worker but rq->curr is the task
actually executing. The former can delay WORKER_CPU_INTENSIVE handling
and pool concurrency management, which can delay pending kernel work
and userspace operations depending on it.

Use rq->curr for the workqueue tick hook while retaining rq->donor for
scheduler accounting.

Fixes: af0c8b2bf67b ("sched: Split scheduler and execution contexts")
Signed-off-by: Hui Su <sh_def@163.com>
---
 kernel/sched/core.c | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index f78275192036..a86fd3d88714 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -5770,8 +5770,8 @@ void sched_tick(void)
 {
 	int cpu = smp_processor_id();
 	struct rq *rq = cpu_rq(cpu);
-	/* accounting goes to the donor task */
-	struct task_struct *donor;
+	/* scheduler accounting goes to the donor task */
+	struct task_struct *curr, *donor;
 	struct rq_flags rf;
 	unsigned long hw_pressure;
 	u64 resched_latency;
@@ -5782,6 +5782,7 @@ void sched_tick(void)
 	sched_clock_tick();
 
 	rq_lock(rq, &rf);
+	curr = rq->curr;
 	donor = rq->donor;
 
 	psi_account_irqtime(rq, donor, NULL);
@@ -5807,8 +5808,8 @@ void sched_tick(void)
 
 	perf_event_task_tick();
 
-	if (donor->flags & PF_WQ_WORKER)
-		wq_worker_tick(donor);
+	if (curr->flags & PF_WQ_WORKER)
+		wq_worker_tick(curr);
 
 	if (!scx_switched_all()) {
 		rq->idle_balance = idle_cpu(cpu);

base-commit: 89a312991dc6e638a36adc43ccb91dbc25504c04
-- 
2.54.0

Re: [PATCH] sched/core: Call wq_worker_tick() for the execution context

[Tejun Heo]

On Wed, Sep 02, 2026 at 11:02:09PM +0800, Hui Su wrote:
> wq_worker_tick() accounts CPU time and detects CPU-intensive work for
> the kworker that is actually running. With proxy execution, rq->donor
> is the scheduling context while rq->curr is the execution context.
> 
> Calling the hook with rq->donor can skip workqueue accounting when a
> kworker is executing on behalf of a donor task. It can also account a
> blocked kworker when the donor is a worker but rq->curr is the task
> actually executing. The former can delay WORKER_CPU_INTENSIVE handling
> and pool concurrency management, which can delay pending kernel work
> and userspace operations depending on it.
> 
> Use rq->curr for the workqueue tick hook while retaining rq->donor for
> scheduler accounting.
> 
> Fixes: af0c8b2bf67b ("sched: Split scheduler and execution contexts")
> Signed-off-by: Hui Su <sh_def@163.com>

Acked-by: Tejun Heo <tj@kernel.org>

Peter, how do you want to route this patch? It can go through either sched
or wq.

Thanks.

-- 
tejun

Re: [PATCH] sched/core: Call wq_worker_tick() for the execution context

[Peter Zijlstra]

On Wed, Sep 02, 2026 at 08:21:38AM -1000, Tejun Heo wrote:
> On Wed, Sep 02, 2026 at 11:02:09PM +0800, Hui Su wrote:
> > wq_worker_tick() accounts CPU time and detects CPU-intensive work for
> > the kworker that is actually running. With proxy execution, rq->donor
> > is the scheduling context while rq->curr is the execution context.
> > 
> > Calling the hook with rq->donor can skip workqueue accounting when a
> > kworker is executing on behalf of a donor task. It can also account a
> > blocked kworker when the donor is a worker but rq->curr is the task
> > actually executing. The former can delay WORKER_CPU_INTENSIVE handling
> > and pool concurrency management, which can delay pending kernel work
> > and userspace operations depending on it.
> > 
> > Use rq->curr for the workqueue tick hook while retaining rq->donor for
> > scheduler accounting.
> > 
> > Fixes: af0c8b2bf67b ("sched: Split scheduler and execution contexts")
> > Signed-off-by: Hui Su <sh_def@163.com>
> 
> Acked-by: Tejun Heo <tj@kernel.org>
> 
> Peter, how do you want to route this patch? It can go through either sched
> or wq.

Sorry, seems this got lost in the email deluge :/ I can take it through
sched/urgent.

[tip: sched/urgent] sched/core: Call wq_worker_tick() for the execution context

[tip-bot2 for Hui Su]

The following commit has been merged into the sched/urgent branch of tip:

Commit-ID:     f5741d2b34519d387edf6e9798fc7030c20a35f3
Gitweb:        https://git.kernel.org/tip/f5741d2b34519d387edf6e9798fc7030c20a35f3
Author:        Hui Su <sh_def@163.com>
AuthorDate:    Wed, 02 Sep 2026 23:02:09 +08:00
Committer:     Peter Zijlstra <peterz@infradead.org>
CommitterDate: Thu, 10 Sep 2026 10:22:52 +02:00

sched/core: Call wq_worker_tick() for the execution context

wq_worker_tick() accounts CPU time and detects CPU-intensive work for
the kworker that is actually running. With proxy execution, rq->donor
is the scheduling context while rq->curr is the execution context.

Calling the hook with rq->donor can skip workqueue accounting when a
kworker is executing on behalf of a donor task. It can also account a
blocked kworker when the donor is a worker but rq->curr is the task
actually executing. The former can delay WORKER_CPU_INTENSIVE handling
and pool concurrency management, which can delay pending kernel work
and userspace operations depending on it.

Use rq->curr for the workqueue tick hook while retaining rq->donor for
scheduler accounting.

Fixes: af0c8b2bf67b ("sched: Split scheduler and execution contexts")
Signed-off-by: Hui Su <sh_def@163.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Tejun Heo <tj@kernel.org>
Link: https://patch.msgid.link/20260902150208.1209922-2-sh_def@163.com
---
 kernel/sched/core.c |  9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index b998ef6..7885ff7 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -5776,8 +5776,8 @@ void sched_tick(void)
 {
 	int cpu = smp_processor_id();
 	struct rq *rq = cpu_rq(cpu);
-	/* accounting goes to the donor task */
-	struct task_struct *donor;
+	/* scheduler accounting goes to the donor task */
+	struct task_struct *curr, *donor;
 	struct rq_flags rf;
 	unsigned long hw_pressure;
 	u64 resched_latency;
@@ -5788,6 +5788,7 @@ void sched_tick(void)
 	sched_clock_tick();
 
 	rq_lock(rq, &rf);
+	curr = rq->curr;
 	donor = rq->donor;
 
 	psi_account_irqtime(rq, donor, NULL);
@@ -5813,8 +5814,8 @@ void sched_tick(void)
 
 	perf_event_task_tick();
 
-	if (donor->flags & PF_WQ_WORKER)
-		wq_worker_tick(donor);
+	if (curr->flags & PF_WQ_WORKER)
+		wq_worker_tick(curr);
 
 	if (!scx_switched_all()) {
 		rq->idle_balance = idle_cpu(cpu);