Re: [PATCH net-next v5 3/6] net: ti: icssg-prueth: Move long delayed work on system_dfl_long_wq

From
Simon Horman <horms@kernel.org>
Date
2026-11-05 20:32:11
Also in
lkml
Reply to
Marco Crivellari, 2026-11-02
On Mon, Nov 02, 2026 at 11:56:53AM +0100, Marco Crivellari wrote:
Currently emac_ndo_open() starts the stats work for the first time with
queue_work(), using system_long_wq. This workqueue should be used when
long works are expected and it is a per-cpu workqueue.

queue_work() on a per-cpu workqueue enqueues the work on the local CPU,
so the first execution of the work is bound to the CPU that opened the
interface.

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")

The stats work handler, icssg_stats_work_handler(), already re-queues
itself on system_dfl_long_wq. Since the work doesn't rely on per-cpu
variables, there is no obvious reason that justify the use of a per-cpu
workqueue for the initial queueing either. So change system_long_wq with
system_dfl_long_wq in emac_ndo_open(), in both icssg_prueth.c and
icssg_prueth_sr1.c, so that the work may benefit from scheduler task
placement.

Cc: MD Danish Anwar <danishanwar@ti.com>
Cc: Roger Quadros <rogerq@kernel.org>
Signed-off-by: Marco Crivellari <redacted>
Reviewed-by: Simon Horman <horms@kernel.org>
Thread (10 messages) 10 messages, 3 authors, last reply 12d ago
Keyboard shortcuts
hback out one level
jnext message in thread / next series-diff position
kprevious message in thread / previous series-diff position
ldrill in
Escclose help / fold thread tree
?toggle this help