From: Lai Jiangshan <jiangshan.ljs@antgroup.com>

To ensure non-reentrancy, __queue_work() attempts to enqueue a work
item to the pool of the currently executing worker. This is not only
unnecessary for an ordered workqueue, where order inherently suggests
non-reentrancy, but it could also disrupt the sequence if the item is
not enqueued on the newest PWQ.

Just queue it to the newest PWQ and let order management guarantees
non-reentrancy.

Fixes: 4c065dbce1e8("workqueue: Enable unbound cpumask update on ordered workqueues")
Signed-off-by: Lai Jiangshan <jiangshan.ljs@antgroup.com>
---
 kernel/workqueue.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/kernel/workqueue.c b/kernel/workqueue.c
index c910f3c28664..d4fecd23ea44 100644
--- a/kernel/workqueue.c
+++ b/kernel/workqueue.c
@@ -2271,9 +2271,13 @@ static void __queue_work(int cpu, struct workqueue_struct *wq,
 	 * If @work was previously on a different pool, it might still be
 	 * running there, in which case the work needs to be queued on that
 	 * pool to guarantee non-reentrancy.
+	 *
+	 * For ordered workqueue, work items must be queued on the newest pwq
+	 * for accurate order management.  Guaranteed order also guarantees
+	 * non-reentrancy.  See the comments above unplug_oldest_pwq().
 	 */
 	last_pool = get_work_pool(work);
-	if (last_pool && last_pool != pool) {
+	if (last_pool && last_pool != pool && !(wq->flags & __WQ_ORDERED)) {
 		struct worker *worker;
 
 		raw_spin_lock(&last_pool->lock);
-- 
2.19.1.6.gb485710b

On 7/3/24 05:27, Lai Jiangshan wrote:
> From: Lai Jiangshan <jiangshan.ljs@antgroup.com>
>
> To ensure non-reentrancy, __queue_work() attempts to enqueue a work
> item to the pool of the currently executing worker. This is not only
> unnecessary for an ordered workqueue, where order inherently suggests
> non-reentrancy, but it could also disrupt the sequence if the item is
> not enqueued on the newest PWQ.
>
> Just queue it to the newest PWQ and let order management guarantees
> non-reentrancy.
>
> Fixes: 4c065dbce1e8("workqueue: Enable unbound cpumask update on ordered workqueues")
> Signed-off-by: Lai Jiangshan <jiangshan.ljs@antgroup.com>
> ---
>   kernel/workqueue.c | 6 +++++-
>   1 file changed, 5 insertions(+), 1 deletion(-)
>
> diff --git a/kernel/workqueue.c b/kernel/workqueue.c
> index c910f3c28664..d4fecd23ea44 100644
> --- a/kernel/workqueue.c
> +++ b/kernel/workqueue.c
> @@ -2271,9 +2271,13 @@ static void __queue_work(int cpu, struct workqueue_struct *wq,
>   	 * If @work was previously on a different pool, it might still be
>   	 * running there, in which case the work needs to be queued on that
>   	 * pool to guarantee non-reentrancy.
> +	 *
> +	 * For ordered workqueue, work items must be queued on the newest pwq
> +	 * for accurate order management.  Guaranteed order also guarantees
> +	 * non-reentrancy.  See the comments above unplug_oldest_pwq().
>   	 */
>   	last_pool = get_work_pool(work);
> -	if (last_pool && last_pool != pool) {
> +	if (last_pool && last_pool != pool && !(wq->flags & __WQ_ORDERED)) {
>   		struct worker *worker;
>   
>   		raw_spin_lock(&last_pool->lock);

Thanks for the fix again.

Acked-by: Waiman Long <longman@redhat.com>

On Wed, Jul 03, 2024 at 05:27:41PM +0800, Lai Jiangshan wrote:
> From: Lai Jiangshan <jiangshan.ljs@antgroup.com>
> 
> To ensure non-reentrancy, __queue_work() attempts to enqueue a work
> item to the pool of the currently executing worker. This is not only
> unnecessary for an ordered workqueue, where order inherently suggests
> non-reentrancy, but it could also disrupt the sequence if the item is
> not enqueued on the newest PWQ.
> 
> Just queue it to the newest PWQ and let order management guarantees
> non-reentrancy.
> 
> Fixes: 4c065dbce1e8("workqueue: Enable unbound cpumask update on ordered workqueues")
> Signed-off-by: Lai Jiangshan <jiangshan.ljs@antgroup.com>

Applied to wq/for-6.10-fixes w/ stable cc added.

Thanks.

-- 
tejun