[RFC PATCH v2] usb: typec: ucsi: Move long delayed work on system_dfl_long_wq

[RFC PATCH v2] usb: typec: ucsi: Move long delayed work on system_dfl_long_wq

[Marco Crivellari]

Currently the code enqueue work items using {queue|mod}_delayed_work(),
using system_long_wq. This workqueue should be used when long works are
expected and it is a per-cpu workqueue.

The function(s) end up calling __queue_delayed_work(), which set a global
timer that could fire anywhere, enqueuing the work where the timer fired.

Unbound works could benefit from scheduler task placement, to optimize
performance and power consumption. Long work shouldn't stick to a single
CPU.

Recently, a new unbound workqueue specific for long running work has
been added:

    c116737e972e ("workqueue: Add system_dfl_long_wq for long unbound works")

Since the workqueue work doesn't rely on per-cpu variables, there is no
obvious reason that justify the use of a per-cpu workqueue. So change
system_long_wq with system_dfl_long_wq so that the work may benefit from
scheduler task placement.

Signed-off-by: Marco Crivellari <marco.crivellari@suse.com>
---
Changes in v2:
- Fixed a wrong change regarding system_long_wq used by queue_work() instead of queue_delayed_work()

Link to v1: https://lore.kernel.org/all/20260507142726.331206-1-marco.crivellari@suse.com/


 drivers/usb/typec/ucsi/ucsi.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/usb/typec/ucsi/ucsi.c b/drivers/usb/typec/ucsi/ucsi.c
index 5b7ad9e99cb9..20fb1fbc5a7a 100644
--- a/drivers/usb/typec/ucsi/ucsi.c
+++ b/drivers/usb/typec/ucsi/ucsi.c
@@ -1985,7 +1985,7 @@ static void ucsi_init_work(struct work_struct *work)
 			return;
 		}
 
-		queue_delayed_work(system_long_wq, &ucsi->work,
+		queue_delayed_work(system_dfl_long_wq, &ucsi->work,
 				   UCSI_ROLE_SWITCH_INTERVAL);
 	}
 }
@@ -2109,7 +2109,7 @@ int ucsi_register(struct ucsi *ucsi)
 		UCSI_BCD_GET_MINOR(ucsi->version),
 		UCSI_BCD_GET_SUBMINOR(ucsi->version));
 
-	queue_delayed_work(system_long_wq, &ucsi->work, 0);
+	queue_delayed_work(system_dfl_long_wq, &ucsi->work, 0);
 
 	ucsi_debugfs_register(ucsi);
 	return 0;
-- 
2.53.0

Re: [RFC PATCH v2] usb: typec: ucsi: Move long delayed work on system_dfl_long_wq

[Marco Crivellari]

On Fri, May 8, 2026 at 4:39 PM Marco Crivellari
<marco.crivellari@suse.com> wrote:
> [...]
>  drivers/usb/typec/ucsi/ucsi.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)

Gentle ping.

Thanks!

-- 

Marco Crivellari

SUSE Labs

Re: [RFC PATCH v2] usb: typec: ucsi: Move long delayed work on system_dfl_long_wq

[Heikki Krogerus]

On Tue, May 19, 2026 at 11:01:18AM +0200, Marco Crivellari wrote:
> On Fri, May 8, 2026 at 4:39 PM Marco Crivellari
> <marco.crivellari@suse.com> wrote:
> > [...]
> >  drivers/usb/typec/ucsi/ucsi.c | 4 ++--
> >  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> Gentle ping.

This is okay by me. FWIW:

Reviewed-by: Heikki Krogerus <heikki.krogerus@linux.intel.com>

thanks,

-- 
heikki

Re: [RFC PATCH v2] usb: typec: ucsi: Move long delayed work on system_dfl_long_wq

[Greg Kroah-Hartman]

On Fri, May 08, 2026 at 04:38:53PM +0200, Marco Crivellari wrote:
> Currently the code enqueue work items using {queue|mod}_delayed_work(),
> using system_long_wq. This workqueue should be used when long works are
> expected and it is a per-cpu workqueue.
> 
> The function(s) end up calling __queue_delayed_work(), which set a global
> timer that could fire anywhere, enqueuing the work where the timer fired.
> 
> Unbound works could benefit from scheduler task placement, to optimize
> performance and power consumption. Long work shouldn't stick to a single
> CPU.
> 
> Recently, a new unbound workqueue specific for long running work has
> been added:
> 
>     c116737e972e ("workqueue: Add system_dfl_long_wq for long unbound works")
> 
> Since the workqueue work doesn't rely on per-cpu variables, there is no
> obvious reason that justify the use of a per-cpu workqueue. So change
> system_long_wq with system_dfl_long_wq so that the work may benefit from
> scheduler task placement.
> 
> Signed-off-by: Marco Crivellari <marco.crivellari@suse.com>
> ---
> Changes in v2:
> - Fixed a wrong change regarding system_long_wq used by queue_work() instead of queue_delayed_work()
> 
> Link to v1: https://lore.kernel.org/all/20260507142726.331206-1-marco.crivellari@suse.com/

What "comments" do you need resolved here before submitting this as a
real patch?

thanks,

greg k-h

Re: [RFC PATCH v2] usb: typec: ucsi: Move long delayed work on system_dfl_long_wq

[Marco Crivellari]

On Fri, May 22, 2026 at 10:37 AM Greg Kroah-Hartman
<gregkh@linuxfoundation.org> wrote:
> [...]
> What "comments" do you need resolved here before submitting this as a
> real patch?
>
> thanks,

Ah, sorry, I forgot to remove the RFC. It is already fine like this.
Should i send a news version removing the RFC prefix or it is not needed?

Thanks!

-- 

Marco Crivellari

SUSE Labs

Re: [RFC PATCH v2] usb: typec: ucsi: Move long delayed work on system_dfl_long_wq

[Greg Kroah-Hartman]

On Fri, May 22, 2026 at 11:27:22AM +0200, Marco Crivellari wrote:
> On Fri, May 22, 2026 at 10:37 AM Greg Kroah-Hartman
> <gregkh@linuxfoundation.org> wrote:
> > [...]
> > What "comments" do you need resolved here before submitting this as a
> > real patch?
> >
> > thanks,
> 
> Ah, sorry, I forgot to remove the RFC. It is already fine like this.
> Should i send a news version removing the RFC prefix or it is not needed?

Please send a new version.