On 17/03/2026 2:35, lirongqing wrote:

From: Li RongQing <redacted>

During device hot-unplug, the mlx5 driver expects quickly unregister
notification chains. The standard atomic_notifier_chain_unregister()
calls synchronize_rcu(), which introduces significant latency and
can become a bottleneck during mass resource cleanup.

Introduce atomic_notifier_chain_unregister_expedited() to leverage
synchronize_rcu_expedited(), and use it significantly reducing wait
times in the following paths:
  - Event Queue (EQ) notifier chain
  - Firmware event notifier chain
  - IRQ notifier chain

On x86-64 with HZ=1000, 64 networking channels:
- Average teardown time: 3.59s -> 1.9s (47% reduction)
On x86-64 with HZ=250, 64 networking channels:
- Average teardown time: 5.5s -> 1.9s (65% reduction)

Co-developed-by: liyongkang <redacted>
Signed-off-by: liyongkang <redacted>
Signed-off-by: Li RongQing <redacted>
---
Diff with v1: fix doc warning and add detailed example

Hi, thanks for your patch.

Direction looks okay overall, nice improvement.
I'd split into 2 patches in a series, though.

This is the kind of patches that should go through our regression tests. 
I'm taking it for testing and will update.

Regards,
Tariq