netdev - Re: [PATCH] net: wwan: iosm: initialize pc_wwan->if

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 PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

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

Date:   Mon, 24 Oct 2022 11:17:09 +0200
From:   Hans de Goede <hdegoede@...hat.com>
To:     Sergey Ryazanov <ryazanov.s.a@...il.com>
Cc:     M Chetan Kumar <m.chetan.kumar@...el.com>,
        Intel Corporation <linuxwwan@...el.com>,
        Loic Poulain <loic.poulain@...aro.org>,
        Johannes Berg <johannes@...solutions.net>,
        netdev@...r.kernel.org
Subject: Re: [PATCH] net: wwan: iosm: initialize pc_wwan->if_mutex earlier

Hi,

On 10/24/22 11:14, Sergey Ryazanov wrote:
> On Mon, Oct 24, 2022 at 12:04 PM Hans de Goede <hdegoede@...hat.com> wrote:
>> On 10/15/22 09:55, Sergey Ryazanov wrote:
>>> On Fri, Oct 14, 2022 at 1:36 PM Hans de Goede <hdegoede@...hat.com> wrote:
>>>> wwan_register_ops() ends up calls ipc_wwan_newlink() before it returns.
>>>>
>>>> ipc_wwan_newlink() uses pc_wwan->if_mutex, so we must initialize it
>>>> before calling wwan_register_ops(). This fixes the following WARN()
>>>> when lock-debugging is enabled:
>>>>
>>>> [skipped]
>>>>
>>>> Signed-off-by: Hans de Goede <hdegoede@...hat.com>
>>>
>>> Should we add a Fixes: tag for this change? Besides this:
>>
>> This issue was present already in the driver as originally introduced, so:
>>
>> Fixes: 2a54f2c77934 ("net: iosm: net driver")
>>
>> I guess?
> 
> The wwan_register_ops() call has been here since the driver
> introduction. But at that time, this call did not create any
> interfaces. The issue was most probably introduced by my change:
> 
> 83068395bbfc ("net: iosm: create default link via WWAN core")
> 
> after which the wwan_register_ops() call was modified in such a way
> that it started calling ipc_wwan_newlink() internally.

Ok, lets go with:

Fixes: 83068395bbfc ("net: iosm: create default link via WWAN core")

Shall I send a v2 with this ? or is this just going to get added
in while merging this?

Regards,

Hans