netdev - Re: [PATCH net 1/3] net/mlx5e: Harden uplink netdev access against device unbind

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: <573a5647-5161-4c4c-bf1f-bc66e8df14a5@nvidia.com>
Date: Mon, 15 Sep 2025 08:59:24 +0800
From: Jianbo Liu <jianbol@...dia.com>
To: Przemek Kitszel <przemyslaw.kitszel@...el.com>
CC: Tariq Toukan <tariqt@...dia.com>, Eric Dumazet <edumazet@...gle.com>,
	Paolo Abeni <pabeni@...hat.com>, Andrew Lunn <andrew+netdev@...n.ch>, "David
 S. Miller" <davem@...emloft.net>, Saeed Mahameed <saeedm@...dia.com>, "Leon
 Romanovsky" <leon@...nel.org>, Mark Bloch <mbloch@...dia.com>,
	<netdev@...r.kernel.org>, <linux-rdma@...r.kernel.org>,
	<linux-kernel@...r.kernel.org>, Gal Pressman <gal@...dia.com>, Jakub Kicinski
	<kuba@...nel.org>
Subject: Re: [PATCH net 1/3] net/mlx5e: Harden uplink netdev access against
 device unbind



On 9/12/2025 7:07 PM, Przemek Kitszel wrote:
> On 9/11/25 09:09, Jianbo Liu wrote:
>>
>>
>> On 9/11/2025 8:45 AM, Jakub Kicinski wrote:
>>> On Wed, 10 Sep 2025 11:23:09 +0800 Jianbo Liu wrote:
>>>> On 9/10/2025 9:23 AM, Jakub Kicinski wrote:
>>>>> On Mon, 8 Sep 2025 13:07:04 +0300 Tariq Toukan wrote:
>>>>>> +    struct net_device *netdev = mlx5_uplink_netdev_get(dev);
>>>>>> +    struct mlx5e_priv *priv;
>>>>>> +    int err;
>>>>>> +
>>>>>> +    if (!netdev)
>>>>>> +        return 0;
>>>>>
>>>>> Please don't call in variable init functions which require cleanup
>>>>> or error checking.
>>>>
>>>> But in this function, a NULL return from mlx5_uplink_netdev_get is a
>>>> valid condition where it should simply return 0. No cleanup or error
>>>> check is needed.
>>>
>>> You have to check if it succeeded, and if so, you need to clean up
>>> later. Do no hide meaningful code in variable init.
>>
>> My focus was on the NULL case, but I see now that the real issue is 
>> ensuring the corresponding cleanup (_put) happens on the successful 
>> path. Hiding the _get call in the initializer makes that less clear.
>>
>> I will refactor the code to follow the correct pattern, like this:
>>
>> struct net_device *netdev;
>>
>> netdev = mlx5_uplink_netdev_get(dev);
>> if (!netdev)
>>      return 0;
>>
>> Thank you for the explanation.
>>
> 
> that would be much better, and make it obvious that there is
> matched get() and put() calls
> 
> would be also great to minify stacktrace
> https://www.kernel.org/doc/html/latest/process/submitting- 
> patches.html#backtraces-in-commit-messages
> 

Got it. I'll minify the stack trace. Thanks for the tip.