linux-kernel - Re: [PATCH] tg3: add new module param to force device power down on reboot

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

Date: Wed, 10 Jan 2024 08:34:18 +0100
From: Heiner Kallweit <hkallweit1@...il.com>
To: Michael Chan <michael.chan@...adcom.com>
Cc: Pavan Chebbi <pavan.chebbi@...adcom.com>,
 Andrea Fois <andrea.fois@...ntsense.it>, Michael Chan <mchan@...adcom.com>,
 "David S. Miller" <davem@...emloft.net>, Eric Dumazet <edumazet@...gle.com>,
 Jakub Kicinski <kuba@...nel.org>, Paolo Abeni <pabeni@...hat.com>,
 George Shuklin <george.shuklin@...il.com>, netdev@...r.kernel.org,
 linux-kernel@...r.kernel.org
Subject: Re: [PATCH] tg3: add new module param to force device power down on
 reboot

On 10.01.2024 08:17, Michael Chan wrote:
> On Tue, Jan 9, 2024 at 11:09 PM Heiner Kallweit <hkallweit1@...il.com> wrote:
>>
>> On 10.01.2024 05:12, Pavan Chebbi wrote:
>>> I think the second suggestion could be a better solution. Helps to
>>> solve the issue 9fc3bc764334 is trying to fix.
>>> But I am not sure how easy it is to test. As I recall, Goerge was
>>> unable to reach out to the author of 2ca1c94ce0b6 when he wanted to
>>> test his patch for regression.
>>> We did discuss the risk of this regression.
>>> https://patchwork.kernel.org/project/netdevbpf/patch/20231101130418.44164-1-george.shuklin@gmail.com/
>>> Unfortunately, looks like it has come true :(
>>
>> If the culprit is an unexpected MSI interrupt, then you may also address
>> this directly by disabling interrupts in the device. tg3_stop() may be
>> a candidate here.
>>
> 
> We already call dev_close() which will call tg3_close() -> tg3_stop()
> a few lines above.

tg3_stop() calls tg3_disable_ints(), so I wonder how a MSI interrupt can
occur after that. Does tg3_disable_ints() disable interrupts synchronously?
Or maybe some kind of commit is needed?