Date: Wed, 23 Nov 2016 19:34:46 +0000
From: Jason Cooper <jason@lakedaemon.net>
To: Kalle Valo <kvalo@codeaurora.org>
Cc: linux-wireless@vger.kernel.org,
        Linux ARM Kernel <linux-arm-kernel@lists.infradead.org>,
        ath9k-devel@qca.qualcomm.com, ath9k-devel@lists.ath9k.org,
        Thomas Petazzoni <thomas.petazzoni@free-electrons.com>,
        Gregory CLEMENT <gregory.clement@free-electrons.com>,
        Andrew Lunn <andrew@lunn.ch>
Subject: Re: ath9k ARMv7 OOPS in v4.8.6, v4.2.8
Message-ID: <20161123193446.GG2799@io.lakedaemon.net> (sfid-20161123_205104_890454_3BC3A29A)
References: <20161123191539.GF2799@io.lakedaemon.net>
 <87a8cqgfpp.fsf@kamboji.qca.qualcomm.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
In-Reply-To: <87a8cqgfpp.fsf@kamboji.qca.qualcomm.com>
Sender: linux-wireless-owner@vger.kernel.org

Hi Kalle,

On Wed, Nov 23, 2016 at 09:26:42PM +0200, Kalle Valo wrote:
> Jason Cooper <jason@lakedaemon.net> writes:
> > I have a Ubiquiti SR-71 mini-pcie ath9k card in a Globalscale Mirabox
> > board (Marvell Armada 370 SoC).  Every day or so I get a consistent
> > crash that brings down the whole board.  I've attached three oops I
> > captured on the serial port.
> >
> > I looked at the commits from v4.8.6 to v4.9-rc6, and nothing jumped out
> > at me as "this would fix it".  And since it takes a day or so to trigger
> > the oops, bisecting would be a bit brutal.  Does anyone have any insight
> > into this?
> 
> Is this a regression, meaning that it didn't crash on older kernels but
> crashes on newer ones? Or has it always crashed?

iirc, it's always done this.  It's one of my spare wifi backhauls that
spends most of it's time in a cardboard box waiting for a task,
collecting dust.  Kinda like the toys in Toy Story.

I pulled it out a month or so ago and the behavior started.  It had
4.2.8 on it at the time.  I upgraded to latest stable a few weeks ago
(v4.8.6) and I'm getting the same issue.

When I originally set it up, it didn't run long enough for me to recall
if the issue occurred.  Best I recall, that was with v4.2.8.

thx,

Jason.