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-prev] [day] [month] [year] [list]
Message-ID: <Z74ekZy_WnOhAQrD@x130>
Date: Tue, 25 Feb 2025 11:48:33 -0800
From: Saeed Mahameed <saeed@...nel.org>
To: Jean-François Roy <jf@...klog.net>
Cc: netdev@...r.kernel.org
Subject: Re: mlx5e_xmit: detected field-spanning write (6.12.16)

On 25 Feb 08:37, Jean-François Roy wrote:
>I'm regularly seeing field-spanning write dumps from the mlx5 driver
>on one of my Talos Linux + Cilium nodes running Linux 6.12.16. I don't
>know if this is caused by a bug in one of Cilum's bpf programs or if
>it's a legitimate issue with the driver.
>
>kantai1: kern: warning: [2025-02-25T16:19:43.741311529Z]:
>------------[ cut here ]------------
>kantai1: kern: warning: [2025-02-25T16:19:43.741322529Z]: memcpy:
>detected field-spanning write (size 32) of single field "h6 + 1" at
>drivers/net/ethernet/mellanox/mlx5/core/en_tx.c:469 (size 0)
>kantai1: kern: warning: [2025-02-25T16:19:43.741350529Z]: WARNING:
>CPU: 2 PID: 5273 at
>drivers/net/ethernet/mellanox/mlx5/core/en_tx.c:469
>mlx5e_xmit+0x99b/0xe00 [mlx5_core]

False alarm 

Can you please test this diff:

  git diff
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c
b/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c
index f8c7912abe0e..40ed9d37edf4 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c
@@ -448,7 +448,7 @@ mlx5e_sq_xmit_wqe(struct mlx5e_txqsq *sq, struct
sk_buff *skb,
         eseg->mss = attr->mss;
  
         if (ihs) {
-               u8 *start = eseg->inline_hdr.start;
+               u8 *start = eseg->inline_hdr.data;
  
                 if (unlikely(attr->hopbyhop)) {
                         /* remove the HBH header.

Thanks,
Saeed.

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ