Re: [PATCH iwl-net] igc: Fix deadlock on module removal

[Kurt Kanzenbach <kurt@linutronix.de>]

[-- Attachment #1: Type: text/plain, Size: 1700 bytes --]

Hi Lukas,

On Sun Apr 14 2024, Lukas Wunner wrote:
> [cc += Roman Lozko who originally reported the issue]
>
> On Sun, Apr 14, 2024 at 09:44:10AM +0200, Kurt Kanzenbach wrote:
>> unregister_netdev() acquires the RNTL lock and releases the LEDs bound
>> to that netdevice. However, netdev_trig_deactivate() and later
>> unregister_netdevice_notifier() try to acquire the RTNL lock again.
>> 
>> Avoid this situation by not using the device-managed LED class
>> functions.
>> 
>> Suggested-by: Lukas Wunner <lukas@wunner.de>
>> Fixes: ea578703b03d ("igc: Add support for LEDs on i225/i226")
>> Signed-off-by: Kurt Kanzenbach <kurt@linutronix.de>
>
> This patch is almost a 1:1 copy of the patch I submitted on April 5:
>
> https://lore.kernel.org/all/ZhBN9p1yOyciXkzw@wunner.de/
>
> I think it is mandatory that you include a Signed-off-by with my name
> in that case.  Arguably the commit author ("From:") should also be me.

I was a bit unsure how to proceed with that. See below.

>
> Moreover this is missing a Reported-by tag with Roman Lozko's name.
>
> AFAICS the only changes that you made are:
> - rename igc_led_teardown() to igc_led_free()
> - rename ret to err
> - replace devm_kcalloc() with kcalloc()
>   (and you introduced a memory leak while doing so, see below)
>
> Honestly I don't see how those small changes justify omitting a
> Signed-off-by or assuming authorship.
>
> I would have been happy to submit a patch myself, I was waiting
> for a Tested-by from Roman or you.

Perfect. I was wondering why you are not submitting the patch
yourself. Then, please go ahead and submit the patch. Feel free to add
my Tested-by.

Thanks,
Kurt

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 861 bytes --]