From mboxrd@z Thu Jan  1 00:00:00 1970
Message-ID: <45378F04.2070500@domain.hid>
Date: Thu, 19 Oct 2006 16:43:16 +0200
From: Wolfgang Grandegger <wg@domain.hid>
MIME-Version: 1.0
Subject: Re: [Xenomai-core] BUG: sleeping function called from invalid context
	at kernel/xenomai/skins/posix/syscall.c:272
References: <45366314.5000904@domain.hid> <45367E43.5020208@domain.hid>
In-Reply-To: <45367E43.5020208@domain.hid>
Content-Type: text/plain; charset=ISO-8859-15; format=flowed
Content-Transfer-Encoding: 7bit
List-Id: "Xenomai life and development \(bug reports, patches,
	discussions\)" <xenomai.xenomai.org>
List-Unsubscribe: <https://mail.gna.org/listinfo/xenomai-core>,
	<mailto:xenomai-core-request@domain.hid>
List-Archive: </public/xenomai-core>
List-Post: <mailto:xenomai@xenomai.org>
List-Help: <mailto:xenomai-core-request@domain.hid>
List-Subscribe: <https://mail.gna.org/listinfo/xenomai-core>,
	<mailto:xenomai-core-request@domain.hid>
To: Jan Kiszka <jan.kiszka@domain.hid>
Cc: xenomai-core <xenomai@xenomai.org>

Jan Kiszka wrote:
> Wolfgang Grandegger wrote:
>> Hello,
>>
>> when I start the RT-Socket-CAN program rtcan_rtt.c, I get the attached
>> error message. Any idea where the problem could be? It was working with
>> Xenomai under Linux 2.4.25.
>>
>> Thanks.
>>
>> Wolfgang.
>>
>>
>>
>> ------------------------------------------------------------------------
>>
>> bash-3.00# rtcan_rtt rtcan0 rtcan1
>> BUG: sleeping function called from invalid context at kernel/xenomai/skins/posix/syscall.c:272
>> in_atomic():0, irqs_disabled():1
> 
> Leaking lock somewhere? Some suggestions:
> 
> - Does it also happen with rtcan_virt?

Yes.

> - If no: switch on XENO_OPT_DEBUG and then XENO_OPT_DEBUG_RTDM (may
>   catch leaking locks of the CAN driver)

Does not give an additional information for the "if yes" case.

> - If yes: what arch are you on? PPC? On x86/2.6 I do not have problems
>   with your demo over rtcan_virt.

It's on PowerPC (MPC5200) under Linux 2.6. I have not realized a similar 
problem with 2.4 on the same board. And it seems to be related with 
pthread_setschedparam(). And is nothing to do with RTCAN as the 
cyclictest() prints the same error message. There is obviously a problem 
with __xn_put_user -> __put_user -> put_user_nocheck:

#define __put_user_nocheck(x, ptr, size)                        \
({                                                              \
         long __pu_err;                                          \
         __typeof__(*(ptr)) __user *__pu_addr = (ptr);           \
         if (!is_kernel_addr((unsigned long)__pu_addr))          \
                 might_sleep();                                  \
         __chk_user_ptr(ptr);                                    \
         __put_user_size((x), __pu_addr, (size), __pu_err);      \
         __pu_err;                                               \
})
 

The "might_sleep" seems to make trouble, which showed up in recent 
versions of Linux on the PowerPC arch :-(.

Wolfgang.



Wolfgang.