netdev - [PATCH net v5 0/2] net: cpsw: Execute ndo_set_rx

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [thread-next>] [day] [month] [year] [list]

Message-Id: <20260203-bbb-v5-0-ea0ea217a85c@gmail.com>
Date: Tue, 03 Feb 2026 10:18:29 +0800
From: Kevin Hao <haokexin@...il.com>
To: netdev@...r.kernel.org
Cc: Kevin Hao <haokexin@...il.com>, stable@...r.kernel.org, 
 Siddharth Vadapalli <s-vadapalli@...com>, Roger Quadros <rogerq@...nel.org>, 
 Andrew Lunn <andrew+netdev@...n.ch>, 
 "David S. Miller" <davem@...emloft.net>, Eric Dumazet <edumazet@...gle.com>, 
 Jakub Kicinski <kuba@...nel.org>, Paolo Abeni <pabeni@...hat.com>, 
 Vladimir Oltean <vladimir.oltean@....com>, 
 Kuniyuki Iwashima <kuniyu@...gle.com>, linux-omap@...r.kernel.org
Subject: [PATCH net v5 0/2] net: cpsw: Execute ndo_set_rx_mode callback in
 a work queue

These two patches resolve an RTNL assertion call trace issue in both the legacy
and new cpsw drivers.

Thanks,
Kevin

---
Changes in v5:
- Apply the same fix to the legacy cpsw driver as well.

- Use goto label in cpsw_ndo_set_rx_mode_work() for the undo path as
  suggested by Jakub.

- Move the disable_work_sync() call to after unregister_netdev() as
  suggested by Jakub.

- Link to v4: https://lore.kernel.org/r/20260130-bbb-v4-1-2bd000a15c34@gmail.com

Changes in v4:
- Using schedule_work() instead of creating a dedicated workqueue.

- Link to v3: https://lore.kernel.org/r/20260127-bbb-v3-1-5e71f340c1e9@gmail.com

Changes in v3:
- Resolve the deadlock issue identified in the AI review [2]
  by moving the netif_running() check under the RTNL lock and removing the
  cancel_work_sync() call in cpsw_ndo_stop().

- Link to v2: https://lore.kernel.org/r/20260125-bbb-v2-1-1547ffabc9d3@gmail.com

Changes in v2:
- Addresses the issue identified in the AI review [1]:
  - Adds a netif_running() check in cpsw_ndo_set_rx_mode_work()
  - Cancels the rx_mode_work in cpsw_ndo_stop()

- Link to v1: https://lore.kernel.org/r/20260123-bbb-v1-1-176b0b71834d@gmail.com

[1] https://netdev-ai.bots.linux.dev/ai-review.html?id=bd885e1e-1aed-4755-ad60-7150737ad0f5
[2] https://netdev-ai.bots.linux.dev/ai-review.html?id=c9fc3cf8-a06c-4cb8-b26b-910e775951a0

---
Kevin Hao (2):
      net: cpsw_new: Execute ndo_set_rx_mode callback in a work queue
      net: cpsw: Execute ndo_set_rx_mode callback in a work queue

 drivers/net/ethernet/ti/cpsw.c      | 41 +++++++++++++++++++++++++++++++------
 drivers/net/ethernet/ti/cpsw_new.c  | 34 +++++++++++++++++++++++++-----
 drivers/net/ethernet/ti/cpsw_priv.h |  1 +
 3 files changed, 65 insertions(+), 11 deletions(-)
---
base-commit: 193579fe01389bc21aff0051d13f24e8ea95b47d
change-id: 20260123-bbb-dc3675f671d0

Best regards,
-- 
Kevin Hao <haokexin@...il.com>