[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <8da9781c-4b89-41cc-8810-8312ef7c2c81@nvidia.com>
Date: Fri, 6 Feb 2026 17:56:08 +0800
From: Jianbo Liu <jianbol@...dia.com>
To: Arnd Bergmann <arnd@...nel.org>, Saeed Mahameed <saeedm@...dia.com>, "Leon
Romanovsky" <leon@...nel.org>, Tariq Toukan <tariqt@...dia.com>, Mark Bloch
<mbloch@...dia.com>, 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>
CC: Arnd Bergmann <arnd@...db.de>, Cosmin Ratiu <cratiu@...dia.com>, "Raed
Salem" <raeds@...dia.com>, <netdev@...r.kernel.org>,
<linux-rdma@...r.kernel.org>, <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] [net-next] net/mlx5e: fix ip6_dst_lookup link failure
On 2/4/2026 9:00 PM, Arnd Bergmann wrote:
> From: Arnd Bergmann <arnd@...db.de>
>
> Changing mlx5 to call ip6_dst_lookup() means it now fails to link
> when IPv6 is a loadable module but ipsec support is built-in:
>
> ipsec.c:(.text+0x1061): undefined reference to `ip6_dst_lookup'
>
> Add a Kconfig dependency that removes avoids this configuration.
>
> Fixes: e35d7da8dd9e ("net/mlx5e: Use ip6_dst_lookup instead of ipv6_dst_lookup_flow for MAC init")
> Signed-off-by: Arnd Bergmann <arnd@...db.de>
> ---
> drivers/net/ethernet/mellanox/mlx5/core/Kconfig | 1 +
> 1 file changed, 1 insertion(+)
>
> diff --git a/drivers/net/ethernet/mellanox/mlx5/core/Kconfig b/drivers/net/ethernet/mellanox/mlx5/core/Kconfig
> index 9cf394c66939..c298efe93f97 100644
> --- a/drivers/net/ethernet/mellanox/mlx5/core/Kconfig
> +++ b/drivers/net/ethernet/mellanox/mlx5/core/Kconfig
> @@ -154,6 +154,7 @@ config MLX5_EN_IPSEC
> depends on MLX5_CORE_EN
> depends on XFRM_OFFLOAD
> depends on INET_ESP_OFFLOAD || INET6_ESP_OFFLOAD
> + depends on IPV6!=m || MLX5_CORE=m
Thanks for the fix.
I received a report for this same error here:
https://lore.kernel.org/oe-kbuild-all/202512261850.P5Jp5BSz-lkp@intel.com/
We were about to send a fix ourselves, it is to simply add:
depends on IPV6 || !IPV6
Is there a specific reason to prefer "depends on IPV6!=m ||
MLX5_CORE=m"? To me, the IPV6 || !IPV6 syntax seems a bit cleaner.
Thanks!
Jianbo
> help
> Build support for IPsec cryptography-offload acceleration in the NIC.
>
Powered by blists - more mailing lists