From mboxrd@z Thu Jan  1 00:00:00 1970
Message-ID: <5076C15E.2050207@gmail.com>
Date: Thu, 11 Oct 2012 14:53:50 +0200
From: Stefan Roese <stefan.roese@gmail.com>
MIME-Version: 1.0
References: <50729DAB.2080909@gmail.com> <50730FB7.3060604@xenomai.org>
	<5073C8C8.7000606@gmail.com> <5074339F.8050601@xenomai.org>
	<50744677.2090103@gmail.com> <5076B407.30709@gmail.com>
	<5076BE30.50605@xenomai.org>
In-Reply-To: <5076BE30.50605@xenomai.org>
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Subject: Re: [Xenomai] Oops while running "cat /proc/xenomai/stat"
List-Id: Discussions about the Xenomai project <xenomai.xenomai.org>
List-Unsubscribe: <http://www.xenomai.org/mailman/options/xenomai>,
	<mailto:xenomai-request@xenomai.org?subject=unsubscribe>
List-Archive: <http://www.xenomai.org/pipermail/xenomai>
List-Post: <mailto:xenomai@xenomai.org>
List-Help: <mailto:xenomai-request@xenomai.org?subject=help>
List-Subscribe: <http://www.xenomai.org/mailman/listinfo/xenomai>,
	<mailto:xenomai-request@xenomai.org?subject=subscribe>
To: Philippe Gerum <rpm@xenomai.org>
Cc: xenomai@xenomai.org

On 10/11/2012 02:40 PM, Philippe Gerum wrote:
> On 10/11/2012 01:56 PM, Stefan Roese wrote:
>> On 10/09/2012 05:44 PM, Stefan Roese wrote:
>>> On 10/09/2012 04:24 PM, Philippe Gerum wrote:
>>>> On 10/09/2012 08:48 AM, Stefan Roese wrote:
>>>>
>>>>> [   65.601569] [c716bba0] [c009adf0] rtdm_event_signal+0x50/0xe4
>>>>> [   65.607440] [c716bbc0] [cb132588] fpga_dma_done_callback+0x18/0x28 [rt_fpga]
>>>>> [   65.614641] [c716bbd0] [cb101114] mpc52xx_lpbfifo_bcom_irq+0x114/0x1c4 [rt_mpc52xx_lpbfifo]
>>>>
>>>> Is any list corruption detected when CONFIG_XENO_OPT_DEBUG_QUEUES is
>>>> enabled?
>>>
>>> No, I don't see any list corruptions with CONFIG_XENO_OPT_DEBUG_QUEUES
>>> enabled. But the crash log is different now. At least part of the
>>> "stat" is printed. And the crash happens now in xnintr_irq_handler().
>>> This different crash seems to result from a change of the kernel
>>> config (I enabled/disabled some other drivers as well in the meantime).
>>> I'm debugging now, how different kernel configurations result in different
>>> crash scenarios.
>>
>> I now strapped down my device driver to the absolute minimum. 
>> "cat /proc/xenomai/stat" still does crash. But not all the time, and not
>> always with the same output. Very strange is the "0x100100" in the
>> output below. This is included in many of the crash reports. Does this
>> ring a bell?
> 
> #define LIST_POISON1  ((void *) 0x00100100 + POISON_POINTER_DELTA)
> 
> Somebody might be doing bad things with memory it does not own anymore?

Yep. That's it. Changing it some other value shows the new changed
value. I'm not doing anything with lists though in my code (AFAIR).

I'll check the increased stack next and let you know.

Thanks,
Stefan