From mboxrd@z Thu Jan  1 00:00:00 1970
Message-ID: <4D725DB2.5030202@domain.hid>
Date: Sat, 05 Mar 2011 16:58:42 +0100
From: Gilles Chanteperdrix <gilles.chanteperdrix@xenomai.org>
MIME-Version: 1.0
References: <AANLkTik6fDir0vM8E_pJerRXY=jvPBtnNnZFO+jwHnOd@domain.hid>
	<4D722CC9.9060307@domain.hid>
In-Reply-To: <4D722CC9.9060307@domain.hid>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 7bit
Subject: Re: [Xenomai-help] Illegal instruction running trivial-periodic on
 ARM	with 2.5.5.2/2.6.33
List-Id: Help regarding installation and common use of Xenomai
	<xenomai.xenomai.org>
List-Unsubscribe: <https://mail.gna.org/options/xenomai-help>,
	<mailto:xenomai-help-request@domain.hid>
List-Archive: </public/xenomai-help>
List-Post: <mailto:xenomai@xenomai.org>
List-Help: <mailto:xenomai-help-request@domain.hid>
List-Subscribe: <https://mail.gna.org/listinfo/xenomai-help>,
	<mailto:xenomai-help-request@domain.hid>
To: Eric Cohen <ericrebates@domain.hid>
Cc: xenomai@xenomai.org

Gilles Chanteperdrix wrote:
> Eric Cohen wrote:
>> Hello, I am attempting to evaluate Xenomai and am running into an
>> illegal instruction when I try to run trivial-periodic.  I am running
>> Xenomai 2.5.5.2 with kernel 2.6.33.5-x3 on a Beagleboard.  Kernel and
>> Xenomai were cross compiled with CodeSourcery 4.4.1.  Upon boot I can
>> see /proc/xenomai/*.  Attempting to run trivial-periodic results in
>> the following output:
>>
>> Xenomai or CONFIG_XENO_OPT_PERVASIVE disabled.
>> (modprobe xeno_nucleus?)
>>
>> Note that CONFIG_XENO_OPT_PERVASIVE is in fact enabled, and nucleus is
>> built into the kernel.  Running in GDB, trivial-periodic receives
>> SIGILL (illegal instruction) and exits.  I get the following
>> backtrace:
>>
>> #0  0x40022498 in xeno_bind_skin_opt () from /usr/xenomai/lib/libxenomai.so.0
>> #1  0x00000000 in ?? ()
>>
>> And the disassembly around the PC is:
>>
>>    0x4002247c <+124>:   movw    r0, #555        ; 0x22b
>>    0x40022480 <+128>:   mov     r6, r5
>>    0x40022484 <+132>:   mov     r1, r8
>>    0x40022488 <+136>:   mov     r2, #1610612743 ; 0x60000007
>>    0x4002248c <+140>:   mov     r3, #3
>>    0x40022490 <+144>:   mov     r4, r7
>>    0x40022494 <+148>:   svc     0x009f0042
>> => 0x40022498 <+152>:   ldr     r12, [sp, #2140]        ; 0x85c
>>    0x4002249c <+156>:   cmp     r12, #0
>>    0x400224a0 <+160>:   beq     0x4002246c <xeno_bind_skin_opt+108>
>>    0x400224a4 <+164>:   mov     r2, r0
>>    0x400224a8 <+168>:   mvn     r1, #84 ; 0x54
>>    0x400224ac <+172>:   mov     r0, r11
>>    0x400224b0 <+176>:   bl      0x400220d8
>>    0x400224b4 <+180>:   cmn     r0, #85 ; 0x55
>>
>> The other example programs result in a SIGILL in exactly the same place.
>>
>> Does anyone have any ideas or suggestions?
> 
> I would say you have incompatible ABI settings for kernel-space and
> user-space.

Here we definitely see an OABI syscall. So, your kernel is probably
configured for EABI.

-- 
                                                                Gilles.