netdev - Re: [net-2.6 PATCH] mlx4_en: Fix a kernel panic when waking tx queue

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]

Message-ID: <4A1B9104.8050405@mellanox.co.il>
Date:	Tue, 26 May 2009 09:49:40 +0300
From:	Yevgeny Petrilin <yevgenyp@...lanox.co.il>
To:	David Miller <davem@...emloft.net>
CC:	netdev@...r.kernel.org, Tziporet Koren <tziporet@...lanox.co.il>
Subject: Re: [net-2.6 PATCH] mlx4_en: Fix a kernel panic when waking tx queue

David Miller wrote:
> From: Yevgeny Petrilin <yevgenyp@...lanox.co.il>
> Date: Mon, 25 May 2009 11:32:42 +0300
> 
>> This patch fixes a kernel panic when waking the TX queue
>> from irq context.
>>
>> Signed-off-by: Yevgeny Petrilin <yevgenyp@...lanox.co.il>
> 
> Why does it fix a panic?  What does the panic look like?
> What causes it?
> 
> Commit message is way too terse, please fix this and resubmit.
> 
Hello Dave,

I did some further investigation on this case.
The patch that I sent was only handling the symptoms instead of fixing the problem.
The panic was caused due to a race when handling TX completions and waking the queue if it was stopped.
When the transmit queue gets full, the driver enables interrupts for TX completions in order to handle
them as fast as possible.
The result was that the driver approached the transmit queue from several different contexts and this
was the bug.
I prepared a new patch fixing the race, it was running for several hours now, the bug didn't happen.

Thanks,
Yevgeny
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html