[PATCH v2 0/2] add WQ_UNBOUND to alloc_workqueue users

[PATCH v2 0/2] add WQ_UNBOUND to alloc_workqueue users

[Marco Crivellari]

Hi,

=== Current situation: problems ===

Let's consider a nohz_full system with isolated CPUs: wq_unbound_cpumask is
set to the housekeeping CPUs, for !WQ_UNBOUND the local CPU is selected.

This leads to different scenarios if a work item is scheduled on an
isolated CPU where "delay" value is 0 or greater then 0:
        schedule_delayed_work(, 0);

This will be handled by __queue_work() that will queue the work item on the
current local (isolated) CPU, while:

        schedule_delayed_work(, 1);

Will move the timer on an housekeeping CPU, and schedule the work there.

Currently if a user enqueue a work item using schedule_delayed_work() the
used wq is "system_wq" (per-cpu wq) while queue_delayed_work() use
WORK_CPU_UNBOUND (used when a cpu is not specified). The same applies to
schedule_work() that is using system_wq and queue_work(), that makes use
again of WORK_CPU_UNBOUND.

This lack of consistency cannot be addressed without refactoring the API.

=== Recent changes to the WQ API ===

The following, address the recent changes in the Workqueue API:

- commit 128ea9f6ccfb ("workqueue: Add system_percpu_wq and system_dfl_wq")
- commit 930c2ea566af ("workqueue: Add new WQ_PERCPU flag")

The old workqueues will be removed in a future release cycle.

=== Introduced Changes by this series ===

1) [P 1-2] add WQ_UNBOUND to alloc_workqueue()

	This explicitly adds WQ_UNBOUND because this workload has no benefit
	from being per-cpu.


Thanks!

---
Changes in v2:
- Switch from WQ_PERCPU to WQ_UNBOUND: this workload benefits from an
  unbound workqueue.

- Rebased on 6.18-rc6


Marco Crivellari (2):
  wifi: rtlwifi: add WQ_UNBOUND to alloc_workqueue users
  wifi: rtw88: add WQ_UNBOUND to alloc_workqueue users

 drivers/net/wireless/realtek/rtlwifi/base.c | 2 +-
 drivers/net/wireless/realtek/rtw88/usb.c    | 3 ++-
 2 files changed, 3 insertions(+), 2 deletions(-)

-- 
2.51.1

[PATCH v2 1/2] wifi: rtlwifi: add WQ_UNBOUND to alloc_workqueue users

[Marco Crivellari]

Currently if a user enqueues a work item using schedule_delayed_work() the
used wq is "system_wq" (per-cpu wq) while queue_delayed_work() use
WORK_CPU_UNBOUND (used when a cpu is not specified). The same applies to
schedule_work() that is using system_wq and queue_work(), that makes use
again of WORK_CPU_UNBOUND.

This lack of consistency cannot be addressed without refactoring the API.
For more details see the Link tag below.

alloc_workqueue() treats all queues as per-CPU by default, while unbound
workqueues must opt-in via WQ_UNBOUND.

This default is suboptimal: most workloads benefit from unbound queues,
allowing the scheduler to place worker threads where they’re needed and
reducing noise when CPUs are isolated.

This continues the effort to refactor workqueue APIs, which began with
the introduction of new workqueues and a new alloc_workqueue flag in:

commit 128ea9f6ccfb ("workqueue: Add system_percpu_wq and system_dfl_wq")
commit 930c2ea566af ("workqueue: Add new WQ_PERCPU flag")

This change adds the WQ_UNBOUND flag to explicitly request
alloc_workqueue() to be unbound, because this specific workload has no
benefit being per-cpu.

With the introduction of the WQ_PERCPU flag (equivalent to !WQ_UNBOUND),
any alloc_workqueue() caller that doesn’t explicitly specify WQ_UNBOUND
must now use WQ_PERCPU.

Once migration is complete, WQ_UNBOUND can be removed and unbound will
become the implicit default.

Suggested-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Marco Crivellari <marco.crivellari@suse.com>
Link: https://lore.kernel.org/all/20250221112003.1dSuoGyc@linutronix.de/
---
 drivers/net/wireless/realtek/rtlwifi/base.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/wireless/realtek/rtlwifi/base.c b/drivers/net/wireless/realtek/rtlwifi/base.c
index e26feb8de658..09e5a16d7252 100644
--- a/drivers/net/wireless/realtek/rtlwifi/base.c
+++ b/drivers/net/wireless/realtek/rtlwifi/base.c
@@ -445,7 +445,7 @@ static int _rtl_init_deferred_work(struct ieee80211_hw *hw)
 	struct rtl_priv *rtlpriv = rtl_priv(hw);
 	struct workqueue_struct *wq;
 
-	wq = alloc_workqueue("%s", 0, 0, rtlpriv->cfg->name);
+	wq = alloc_workqueue("%s", WQ_UNBOUND, 0, rtlpriv->cfg->name);
 	if (!wq)
 		return -ENOMEM;
 
-- 
2.51.1

Re: [PATCH v2 1/2] wifi: rtlwifi: add WQ_UNBOUND to alloc_workqueue users

[Ping-Ke Shih]

Marco Crivellari <marco.crivellari@suse.com> wrote:

> Currently if a user enqueues a work item using schedule_delayed_work() the
> used wq is "system_wq" (per-cpu wq) while queue_delayed_work() use
> WORK_CPU_UNBOUND (used when a cpu is not specified). The same applies to
> schedule_work() that is using system_wq and queue_work(), that makes use
> again of WORK_CPU_UNBOUND.
> 
> This lack of consistency cannot be addressed without refactoring the API.
> For more details see the Link tag below.
> 
> alloc_workqueue() treats all queues as per-CPU by default, while unbound
> workqueues must opt-in via WQ_UNBOUND.
> 
> This default is suboptimal: most workloads benefit from unbound queues,
> allowing the scheduler to place worker threads where they’re needed and
> reducing noise when CPUs are isolated.
> 
> This continues the effort to refactor workqueue APIs, which began with
> the introduction of new workqueues and a new alloc_workqueue flag in:
> 
> commit 128ea9f6ccfb ("workqueue: Add system_percpu_wq and system_dfl_wq")
> commit 930c2ea566af ("workqueue: Add new WQ_PERCPU flag")
> 
> This change adds the WQ_UNBOUND flag to explicitly request
> alloc_workqueue() to be unbound, because this specific workload has no
> benefit being per-cpu.
> 
> With the introduction of the WQ_PERCPU flag (equivalent to !WQ_UNBOUND),
> any alloc_workqueue() caller that doesn’t explicitly specify WQ_UNBOUND
> must now use WQ_PERCPU.
> 
> Once migration is complete, WQ_UNBOUND can be removed and unbound will
> become the implicit default.
> 
> Suggested-by: Tejun Heo <tj@kernel.org>
> Signed-off-by: Marco Crivellari <marco.crivellari@suse.com>
> Link: https://lore.kernel.org/all/20250221112003.1dSuoGyc@linutronix.de/

2 patch(es) applied to rtw-next branch of rtw.git, thanks.

7cbec00dc731 wifi: rtlwifi: add WQ_UNBOUND to alloc_workqueue users
9c194fe4625d wifi: rtw88: add WQ_UNBOUND to alloc_workqueue users

---
https://github.com/pkshih/rtw.git

Re: [PATCH v2 1/2] wifi: rtlwifi: add WQ_UNBOUND to alloc_workqueue users

[Marco Crivellari]

On Fri, Nov 21, 2025 at 4:49 AM Ping-Ke Shih <pkshih@realtek.com> wrote:
>[...]
> 2 patch(es) applied to rtw-next branch of rtw.git, thanks.
>
> 7cbec00dc731 wifi: rtlwifi: add WQ_UNBOUND to alloc_workqueue users
> 9c194fe4625d wifi: rtw88: add WQ_UNBOUND to alloc_workqueue users

Many thanks!

-- 

Marco Crivellari

L3 Support Engineer, Technology & Product

[PATCH v2 2/2] wifi: rtw88: add WQ_UNBOUND to alloc_workqueue users

[Marco Crivellari]

Currently if a user enqueues a work item using schedule_delayed_work() the
used wq is "system_wq" (per-cpu wq) while queue_delayed_work() use
WORK_CPU_UNBOUND (used when a cpu is not specified). The same applies to
schedule_work() that is using system_wq and queue_work(), that makes use
again of WORK_CPU_UNBOUND.

This lack of consistency cannot be addressed without refactoring the API.
For more details see the Link tag below.

alloc_workqueue() treats all queues as per-CPU by default, while unbound
workqueues must opt-in via WQ_UNBOUND.

This default is suboptimal: most workloads benefit from unbound queues,
allowing the scheduler to place worker threads where they’re needed and
reducing noise when CPUs are isolated.

This continues the effort to refactor workqueue APIs, which began with
the introduction of new workqueues and a new alloc_workqueue flag in:

commit 128ea9f6ccfb ("workqueue: Add system_percpu_wq and system_dfl_wq")
commit 930c2ea566af ("workqueue: Add new WQ_PERCPU flag")

This change adds the WQ_UNBOUND flag to explicitly request
alloc_workqueue() to be unbound, because this specific workload has no
benefit being per-cpu.

With the introduction of the WQ_PERCPU flag (equivalent to !WQ_UNBOUND),
any alloc_workqueue() caller that doesn’t explicitly specify WQ_UNBOUND
must now use WQ_PERCPU.

Once migration is complete, WQ_UNBOUND can be removed and unbound will
become the implicit default.

Suggested-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Marco Crivellari <marco.crivellari@suse.com>
Link: https://lore.kernel.org/all/20250221112003.1dSuoGyc@linutronix.de/
---
 drivers/net/wireless/realtek/rtw88/usb.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/net/wireless/realtek/rtw88/usb.c b/drivers/net/wireless/realtek/rtw88/usb.c
index 3b5126ffc81a..009202c627d2 100644
--- a/drivers/net/wireless/realtek/rtw88/usb.c
+++ b/drivers/net/wireless/realtek/rtw88/usb.c
@@ -965,7 +965,8 @@ static int rtw_usb_init_rx(struct rtw_dev *rtwdev)
 	struct sk_buff *rx_skb;
 	int i;
 
-	rtwusb->rxwq = alloc_workqueue("rtw88_usb: rx wq", WQ_BH, 0);
+	rtwusb->rxwq = alloc_workqueue("rtw88_usb: rx wq", WQ_BH | WQ_UNBOUND,
+				       0);
 	if (!rtwusb->rxwq) {
 		rtw_err(rtwdev, "failed to create RX work queue\n");
 		return -ENOMEM;
-- 
2.51.1