All of lore.kernel.org
 help / color / mirror / Atom feed
* [Xenomai] Peak PCI CAN card RTCAN problem
@ 2012-06-23 14:24 Kronander Klas
  2012-06-23 17:49 ` Gilles Chanteperdrix
  0 siblings, 1 reply; 17+ messages in thread
From: Kronander Klas @ 2012-06-23 14:24 UTC (permalink / raw)
  To: xenomai

Hi,

I have an installation of ubuntu 12.04 with xenomai 2.5.5.2. This PC is used to communicate with a robot over CAN bus. The CAN card installed on the PC is a dual channel PCI card from PEAK Systems.

I have not been able to use the CAN port at all so far. The device is detected by the systems and the interfaces show up:

cat /proc/rtcan/devices
Name___________ _Baudrate State___ TX_Counter RX_Counter ____Errors
rtcan0          undefined stopped           0          0          0
rtcan1          undefined stopped           0          0          0

but I am unable to change the state from "stopped" to something useful. When I try to reinitialize it tells me that the driver is not loaded:
/usr/xenomai/sbin/rtcanconfig rtcan0 -b 1000000 -c none start
Cannot open RTDM CAN socket. Maybe driver not loaded?

This is confirmed when looking at the loaded modules with lsmod | grep xeno_can_peak_pci, which does not find the relevant module. When I try to load the module with modprobe xeno_can_peak_pci I do not get any error messages, but the module still does not show up when doing lsmod | grep xeno_can_peak_pci.

Looking in the syslog, I can find some information regarding the CAN device, which seemed to have initialized during system startup:

Jun 23 16:20:37 wam-desktop kernel: [    2.105390] RT-Socket-CAN 0.90.2 - (C) 2006 RT-Socket-CAN Development Team
Jun 23 16:20:37 wam-desktop kernel: [    2.105397] RTCAN SJA1000 driver initialized
Jun 23 16:20:37 wam-desktop kernel: [    2.105405] PEAK-PCI-CAN: initializing device 001c:0001
Jun 23 16:20:37 wam-desktop kernel: [    2.105411] PEAK-PCI-CAN 0000:05:05.0: PCI INT A -> GSI 17 (level, low) -> IRQ 17
Jun 23 16:20:37 wam-desktop kernel: [    2.110385] PEAK-PCI-CAN: base_addr=f8272000 conf_addr=f826e000 irq=17
Jun 23 16:20:37 wam-desktop kernel: [    2.110398] rtcan: registered rtcan0
Jun 23 16:20:37 wam-desktop kernel: [    2.110403] PEAK-PCI-CAN: base_addr=f8276400 conf_addr=f826e000 irq=17
Jun 23 16:20:37 wam-desktop kernel: [    2.110413] rtcan: registered rtcan1

This looks fine right? Yet no driver related to any CAN device is loaded. I am stuck and would greatly appreciate some help for fixing this problem.

Thanks in advance,

Best Regards,
Klas Kronander



^ permalink raw reply	[flat|nested] 17+ messages in thread

* Re: [Xenomai] Peak PCI CAN card RTCAN problem
  2012-06-23 14:24 [Xenomai] Peak PCI CAN card RTCAN problem Kronander Klas
@ 2012-06-23 17:49 ` Gilles Chanteperdrix
  2012-06-28 11:11   ` Kronander Klas
  0 siblings, 1 reply; 17+ messages in thread
From: Gilles Chanteperdrix @ 2012-06-23 17:49 UTC (permalink / raw)
  To: Kronander Klas; +Cc: xenomai

On 06/23/2012 04:24 PM, Kronander Klas wrote:
> Hi,
> 
> I have an installation of ubuntu 12.04 with xenomai 2.5.5.2.

Do you have the same issues with xenomai 2.5.6, or even better, xenomai
2.6.0 ?

-- 
                                                                Gilles.


^ permalink raw reply	[flat|nested] 17+ messages in thread

* Re: [Xenomai] Peak PCI CAN card RTCAN problem
  2012-06-23 17:49 ` Gilles Chanteperdrix
@ 2012-06-28 11:11   ` Kronander Klas
  2012-06-28 11:28     ` Wolfgang Grandegger
  2012-06-28 12:45     ` Gilles Chanteperdrix
  0 siblings, 2 replies; 17+ messages in thread
From: Kronander Klas @ 2012-06-28 11:11 UTC (permalink / raw)
  To: Gilles Chanteperdrix; +Cc: xenomai

Hi,

Thanks for the response. I now tried with xenomai version 2.6, I have some progress but it is still not working. 

CAN-problem:
Looking at the syslog, the can card is detected and the rtcan interfaces are set up properly:

Jun 28 11:58:14 wam-desktop kernel: [   11.954476] RT-Socket-CAN 0.90.2 - (C) 2006 RT-Socket-CAN Development Team
Jun 28 11:58:14 wam-desktop kernel: [   11.960228] RTCAN SJA1000 driver initialized
Jun 28 11:58:14 wam-desktop kernel: [   11.967911] PEAK-PCI-CAN: initializing device 001c:0001
Jun 28 11:58:14 wam-desktop kernel: [   11.967920] PEAK-PCI-CAN 0000:05:05.0: PCI INT A -> GSI 17 (level, low) -> IRQ 17
Jun 28 11:58:14 wam-desktop kernel: [   11.972902] PEAK-PCI-CAN: base_addr=f8470000 conf_addr=f8466000 irq=17
Jun 28 11:58:14 wam-desktop kernel: [   11.972914] rtcan: registered rtcan0
Jun 28 11:58:14 wam-desktop kernel: [   11.972920] PEAK-PCI-CAN: base_addr=f849e400 conf_addr=f8466000 irq=17
Jun 28 11:58:14 wam-desktop kernel: [   11.972928] rtcan: registered rtcan1

I now also see the loaded xeno_can_peak_pci driver with lsmod. The status of the can is STOPPED for both ports:
cat /proc/rtcan/devices
Name___________ _Baudrate State___ TX_Counter RX_Counter ____Errors
rtcan0          undefined stopped           0          0          0
rtcan1          undefined stopped           0          0          0

When trying to start rtcan0 I get an error:
/usr/xenomai/sbin/rtcanconfig rtcan0 -b 1000000 -c none start -v 
Cannot open RTDM CAN socket. Maybe driver not loaded?

But the driver IS loaded, so there is another problem. I looked in the syslog and could found this line:

Jun 28 11:58:14 wam-desktop kernel: [   13.017820] Xenomai: xnintr_shirq_handler: IRQ17 not handled. Disabling IRQ line.

I don't know exactly what happens there or why, but it seems that IRQ 17 is disabled? Why and when does this happen?

On IRQ 17 there is the CAN card and two usb controllers. In the rc.local we try to disable the conflicting devices using:

echo -n "0000:00:1a.1" > /sys/bus/pci/drivers/uhci_hcd/unbind
echo -n "0000:00:1d.1" > /sys/bus/pci/drivers/uhci_hcd/unbind

But that does not make any difference. Thankful for any ideas or suggestions on how to proceed!

Thanks in advance,
Klas Kronander






________________________________________
From: Gilles Chanteperdrix [gilles.chanteperdrix@xenomai.org]
Sent: 23 June 2012 19:49
To: Kronander Klas
Cc: xenomai@xenomai.org
Subject: Re: [Xenomai] Peak PCI CAN card RTCAN problem

On 06/23/2012 04:24 PM, Kronander Klas wrote:
> Hi,
>
> I have an installation of ubuntu 12.04 with xenomai 2.5.5.2.

Do you have the same issues with xenomai 2.5.6, or even better, xenomai
2.6.0 ?

--
                                                                Gilles.


^ permalink raw reply	[flat|nested] 17+ messages in thread

* Re: [Xenomai] Peak PCI CAN card RTCAN problem
  2012-06-28 11:11   ` Kronander Klas
@ 2012-06-28 11:28     ` Wolfgang Grandegger
  2012-06-28 11:38       ` Wolfgang Grandegger
  2012-06-28 12:45     ` Gilles Chanteperdrix
  1 sibling, 1 reply; 17+ messages in thread
From: Wolfgang Grandegger @ 2012-06-28 11:28 UTC (permalink / raw)
  To: Kronander Klas; +Cc: xenomai

On 06/28/2012 01:11 PM, Kronander Klas wrote:
> Hi,
> 
> Thanks for the response. I now tried with xenomai version 2.6, I have some progress but it is still not working. 
> 
> CAN-problem:
> Looking at the syslog, the can card is detected and the rtcan interfaces are set up properly:
> 
> Jun 28 11:58:14 wam-desktop kernel: [   11.954476] RT-Socket-CAN 0.90.2 - (C) 2006 RT-Socket-CAN Development Team
> Jun 28 11:58:14 wam-desktop kernel: [   11.960228] RTCAN SJA1000 driver initialized
> Jun 28 11:58:14 wam-desktop kernel: [   11.967911] PEAK-PCI-CAN: initializing device 001c:0001
> Jun 28 11:58:14 wam-desktop kernel: [   11.967920] PEAK-PCI-CAN 0000:05:05.0: PCI INT A -> GSI 17 (level, low) -> IRQ 17
> Jun 28 11:58:14 wam-desktop kernel: [   11.972902] PEAK-PCI-CAN: base_addr=f8470000 conf_addr=f8466000 irq=17
> Jun 28 11:58:14 wam-desktop kernel: [   11.972914] rtcan: registered rtcan0
> Jun 28 11:58:14 wam-desktop kernel: [   11.972920] PEAK-PCI-CAN: base_addr=f849e400 conf_addr=f8466000 irq=17
> Jun 28 11:58:14 wam-desktop kernel: [   11.972928] rtcan: registered rtcan1
> 
> I now also see the loaded xeno_can_peak_pci driver with lsmod. The status of the can is STOPPED for both ports:
> cat /proc/rtcan/devices
> Name___________ _Baudrate State___ TX_Counter RX_Counter ____Errors
> rtcan0          undefined stopped           0          0          0
> rtcan1          undefined stopped           0          0          0
> 
> When trying to start rtcan0 I get an error:
> /usr/xenomai/sbin/rtcanconfig rtcan0 -b 1000000 -c none start -v 
> Cannot open RTDM CAN socket. Maybe driver not loaded?

Does the following command work?

/usr/xenomai/sbin/rtcanconfig -b 1000000 -c none start -v rtcan0




^ permalink raw reply	[flat|nested] 17+ messages in thread

* Re: [Xenomai] Peak PCI CAN card RTCAN problem
  2012-06-28 11:28     ` Wolfgang Grandegger
@ 2012-06-28 11:38       ` Wolfgang Grandegger
  2012-06-28 11:45         ` Kronander Klas
  0 siblings, 1 reply; 17+ messages in thread
From: Wolfgang Grandegger @ 2012-06-28 11:38 UTC (permalink / raw)
  To: Kronander Klas; +Cc: xenomai

On 06/28/2012 01:28 PM, Wolfgang Grandegger wrote:
> On 06/28/2012 01:11 PM, Kronander Klas wrote:
>> Hi,
>>
>> Thanks for the response. I now tried with xenomai version 2.6, I have some progress but it is still not working. 
>>
>> CAN-problem:
>> Looking at the syslog, the can card is detected and the rtcan interfaces are set up properly:
>>
>> Jun 28 11:58:14 wam-desktop kernel: [   11.954476] RT-Socket-CAN 0.90.2 - (C) 2006 RT-Socket-CAN Development Team
>> Jun 28 11:58:14 wam-desktop kernel: [   11.960228] RTCAN SJA1000 driver initialized
>> Jun 28 11:58:14 wam-desktop kernel: [   11.967911] PEAK-PCI-CAN: initializing device 001c:0001
>> Jun 28 11:58:14 wam-desktop kernel: [   11.967920] PEAK-PCI-CAN 0000:05:05.0: PCI INT A -> GSI 17 (level, low) -> IRQ 17
>> Jun 28 11:58:14 wam-desktop kernel: [   11.972902] PEAK-PCI-CAN: base_addr=f8470000 conf_addr=f8466000 irq=17
>> Jun 28 11:58:14 wam-desktop kernel: [   11.972914] rtcan: registered rtcan0
>> Jun 28 11:58:14 wam-desktop kernel: [   11.972920] PEAK-PCI-CAN: base_addr=f849e400 conf_addr=f8466000 irq=17
>> Jun 28 11:58:14 wam-desktop kernel: [   11.972928] rtcan: registered rtcan1
>>
>> I now also see the loaded xeno_can_peak_pci driver with lsmod. The status of the can is STOPPED for both ports:
>> cat /proc/rtcan/devices
>> Name___________ _Baudrate State___ TX_Counter RX_Counter ____Errors
>> rtcan0          undefined stopped           0          0          0
>> rtcan1          undefined stopped           0          0          0
>>
>> When trying to start rtcan0 I get an error:
>> /usr/xenomai/sbin/rtcanconfig rtcan0 -b 1000000 -c none start -v 
>> Cannot open RTDM CAN socket. Maybe driver not loaded?
> 
> Does the following command work?
> 
> /usr/xenomai/sbin/rtcanconfig -b 1000000 -c none start -v rtcan0

Sorry, it should be

Usage: rtcanconfig <can-interface> [Options] [up|down|start|stop|sleep]

Therefore try:

#/usr/xenomai/sbin/rtcanconfig rtcan0 -b 1000000 -v start

Wolfgang


^ permalink raw reply	[flat|nested] 17+ messages in thread

* Re: [Xenomai] Peak PCI CAN card RTCAN problem
  2012-06-28 11:38       ` Wolfgang Grandegger
@ 2012-06-28 11:45         ` Kronander Klas
  2012-06-28 11:57           ` Wolfgang Grandegger
  0 siblings, 1 reply; 17+ messages in thread
From: Kronander Klas @ 2012-06-28 11:45 UTC (permalink / raw)
  To: Wolfgang Grandegger; +Cc: xenomai

Hi,

I still get the same error:

mode: start(0x1)
Cannot open RTDM CAN Socket. Maybe Driver not loaded? 

I guess this error message is related to this line in my syslog (?) :

Jun 28 11:58:14 wam-desktop kernel: [   13.017820] Xenomai: xnintr_shirq_handler: IRQ17 not handled. Disabling IRQ line. 

the CAN card is on IRQ 17. 

Thanks,
Klas Kronander
________________________________________
From: Wolfgang Grandegger [wg@grandegger.com]
Sent: 28 June 2012 13:38
To: Kronander Klas
Cc: xenomai@xenomai.org
Subject: Re: [Xenomai] Peak PCI CAN card RTCAN problem

On 06/28/2012 01:28 PM, Wolfgang Grandegger wrote:
> On 06/28/2012 01:11 PM, Kronander Klas wrote:
>> Hi,
>>
>> Thanks for the response. I now tried with xenomai version 2.6, I have some progress but it is still not working.
>>
>> CAN-problem:
>> Looking at the syslog, the can card is detected and the rtcan interfaces are set up properly:
>>
>> Jun 28 11:58:14 wam-desktop kernel: [   11.954476] RT-Socket-CAN 0.90.2 - (C) 2006 RT-Socket-CAN Development Team
>> Jun 28 11:58:14 wam-desktop kernel: [   11.960228] RTCAN SJA1000 driver initialized
>> Jun 28 11:58:14 wam-desktop kernel: [   11.967911] PEAK-PCI-CAN: initializing device 001c:0001
>> Jun 28 11:58:14 wam-desktop kernel: [   11.967920] PEAK-PCI-CAN 0000:05:05.0: PCI INT A -> GSI 17 (level, low) -> IRQ 17
>> Jun 28 11:58:14 wam-desktop kernel: [   11.972902] PEAK-PCI-CAN: base_addr=f8470000 conf_addr=f8466000 irq=17
>> Jun 28 11:58:14 wam-desktop kernel: [   11.972914] rtcan: registered rtcan0
>> Jun 28 11:58:14 wam-desktop kernel: [   11.972920] PEAK-PCI-CAN: base_addr=f849e400 conf_addr=f8466000 irq=17
>> Jun 28 11:58:14 wam-desktop kernel: [   11.972928] rtcan: registered rtcan1
>>
>> I now also see the loaded xeno_can_peak_pci driver with lsmod. The status of the can is STOPPED for both ports:
>> cat /proc/rtcan/devices
>> Name___________ _Baudrate State___ TX_Counter RX_Counter ____Errors
>> rtcan0          undefined stopped           0          0          0
>> rtcan1          undefined stopped           0          0          0
>>
>> When trying to start rtcan0 I get an error:
>> /usr/xenomai/sbin/rtcanconfig rtcan0 -b 1000000 -c none start -v
>> Cannot open RTDM CAN socket. Maybe driver not loaded?
>
> Does the following command work?
>
> /usr/xenomai/sbin/rtcanconfig -b 1000000 -c none start -v rtcan0

Sorry, it should be

Usage: rtcanconfig <can-interface> [Options] [up|down|start|stop|sleep]

Therefore try:

#/usr/xenomai/sbin/rtcanconfig rtcan0 -b 1000000 -v start

Wolfgang


^ permalink raw reply	[flat|nested] 17+ messages in thread

* Re: [Xenomai] Peak PCI CAN card RTCAN problem
  2012-06-28 11:45         ` Kronander Klas
@ 2012-06-28 11:57           ` Wolfgang Grandegger
  2012-06-28 12:52             ` Klas Kronander
  0 siblings, 1 reply; 17+ messages in thread
From: Wolfgang Grandegger @ 2012-06-28 11:57 UTC (permalink / raw)
  To: Kronander Klas; +Cc: xenomai

Hello,

On 06/28/2012 01:45 PM, Kronander Klas wrote:
> Hi,
> 
> I still get the same error:
> 
> mode: start(0x1)
> Cannot open RTDM CAN Socket. Maybe Driver not loaded? 
> 
> I guess this error message is related to this line in my syslog (?) :
> 
> Jun 28 11:58:14 wam-desktop kernel: [   13.017820] Xenomai: xnintr_shirq_handler: IRQ17 not handled. Disabling IRQ line. 
> 
> the CAN card is on IRQ 17. 

I don't think so. It's an unrelated issue. There is a problem with the
protocol in the first place. Could you please hack "rtcanconfig.c" to
print the return value of rt_dev_socket? Most likely it will return
-EPROTONOSUPPORT.

Wolfgang.


^ permalink raw reply	[flat|nested] 17+ messages in thread

* Re: [Xenomai] Peak PCI CAN card RTCAN problem
  2012-06-28 11:11   ` Kronander Klas
  2012-06-28 11:28     ` Wolfgang Grandegger
@ 2012-06-28 12:45     ` Gilles Chanteperdrix
  1 sibling, 0 replies; 17+ messages in thread
From: Gilles Chanteperdrix @ 2012-06-28 12:45 UTC (permalink / raw)
  To: Kronander Klas; +Cc: xenomai

On 06/28/2012 01:11 PM, Kronander Klas wrote:
> Jun 28 11:58:14 wam-desktop kernel: [   13.017820] Xenomai:
> xnintr_shirq_handler: IRQ17 not handled. Disabling IRQ line.
> 
> I don't know exactly what happens there or why, but it seems that IRQ
> 17 is disabled? Why and when does this happen?
> 
> On IRQ 17 there is the CAN card and two usb controllers. In the
> rc.local we try to disable the conflicting devices using:

That is a problem (maybe not what causes the issues you have, but 
something you will have to deal with eventually), see:

http://xenomai.org/index.php/FAQs#What_can_I_do_if_Xenomai_and_Linux_devices_share_the_same_IRQ.3F

-- 
					    Gilles.


^ permalink raw reply	[flat|nested] 17+ messages in thread

* Re: [Xenomai] Peak PCI CAN card RTCAN problem
  2012-06-28 11:57           ` Wolfgang Grandegger
@ 2012-06-28 12:52             ` Klas Kronander
  2012-06-28 13:07               ` Wolfgang Grandegger
  0 siblings, 1 reply; 17+ messages in thread
From: Klas Kronander @ 2012-06-28 12:52 UTC (permalink / raw)
  To: Wolfgang Grandegger; +Cc: xenomai

Hi,

ok did that. The return value from rt_dev_socket is -14. I could not 
find the definition of EPROTONOSUPPORT anywhere, so I don't know if this 
was the expected value?

Thanks,
Klas


On 2012-06-28 13:57, Wolfgang Grandegger wrote:
> Hello,
>
> On 06/28/2012 01:45 PM, Kronander Klas wrote:
>> Hi,
>>
>> I still get the same error:
>>
>> mode: start(0x1)
>> Cannot open RTDM CAN Socket. Maybe Driver not loaded?
>>
>> I guess this error message is related to this line in my syslog (?) :
>>
>> Jun 28 11:58:14 wam-desktop kernel: [   13.017820] Xenomai: xnintr_shirq_handler: IRQ17 not handled. Disabling IRQ line.
>>
>> the CAN card is on IRQ 17.
> I don't think so. It's an unrelated issue. There is a problem with the
> protocol in the first place. Could you please hack "rtcanconfig.c" to
> print the return value of rt_dev_socket? Most likely it will return
> -EPROTONOSUPPORT.
>
> Wolfgang.



^ permalink raw reply	[flat|nested] 17+ messages in thread

* Re: [Xenomai] Peak PCI CAN card RTCAN problem
  2012-06-28 12:52             ` Klas Kronander
@ 2012-06-28 13:07               ` Wolfgang Grandegger
  2012-06-28 13:27                 ` Wolfgang Grandegger
  0 siblings, 1 reply; 17+ messages in thread
From: Wolfgang Grandegger @ 2012-06-28 13:07 UTC (permalink / raw)
  To: Klas Kronander; +Cc: xenomai

On 06/28/2012 02:52 PM, Klas Kronander wrote:
> Hi,
> 
> ok did that. The return value from rt_dev_socket is -14. I could not
> find the definition of EPROTONOSUPPORT anywhere, so I don't know if this
> was the expected value?

~$ errno 14
#define	EFAULT		14	/* Bad address */

Hm, strange. I have to look that up.
Wolfgang.


^ permalink raw reply	[flat|nested] 17+ messages in thread

* Re: [Xenomai] Peak PCI CAN card RTCAN problem
  2012-06-28 13:07               ` Wolfgang Grandegger
@ 2012-06-28 13:27                 ` Wolfgang Grandegger
  2012-06-28 16:09                   ` Klas Kronander
  0 siblings, 1 reply; 17+ messages in thread
From: Wolfgang Grandegger @ 2012-06-28 13:27 UTC (permalink / raw)
  To: Klas Kronander; +Cc: xenomai

On 06/28/2012 03:07 PM, Wolfgang Grandegger wrote:
> On 06/28/2012 02:52 PM, Klas Kronander wrote:
>> Hi,
>>
>> ok did that. The return value from rt_dev_socket is -14. I could not
>> find the definition of EPROTONOSUPPORT anywhere, so I don't know if this
>> was the expected value?
> 
> ~$ errno 14
> #define	EFAULT		14	/* Bad address */
> 
> Hm, strange. I have to look that up.

To better understand what you are doing... are you using a Xenomai
enabled kernel? Does the latency test work?

Wolfgang.


^ permalink raw reply	[flat|nested] 17+ messages in thread

* Re: [Xenomai] Peak PCI CAN card RTCAN problem
  2012-06-28 13:27                 ` Wolfgang Grandegger
@ 2012-06-28 16:09                   ` Klas Kronander
  2012-06-28 16:12                     ` Gilles Chanteperdrix
  0 siblings, 1 reply; 17+ messages in thread
From: Klas Kronander @ 2012-06-28 16:09 UTC (permalink / raw)
  To: Wolfgang Grandegger; +Cc: xenomai



On 2012-06-28 15:27, Wolfgang Grandegger wrote:
> On 06/28/2012 03:07 PM, Wolfgang Grandegger wrote:
>> On 06/28/2012 02:52 PM, Klas Kronander wrote:
>>> Hi,
>>>
>>> ok did that. The return value from rt_dev_socket is -14. I could not
>>> find the definition of EPROTONOSUPPORT anywhere, so I don't know if this
>>> was the expected value?
>> ~$ errno 14
>> #define	EFAULT		14	/* Bad address */
>>
>> Hm, strange. I have to look that up.
> To better understand what you are doing... are you using a Xenomai
> enabled kernel? Does the latency test work?
>
> Wolfgang.

Yes, I am using a Xenomai enabled kernel. None of the tests seem to work 
though.

clocktest just gives me zeros. latency does not show anything. I have 
tried to recompile the kernel, still the same result. How find out 
what's wrong? I followed the guidlines on the xenomai website for the 
kernel configuration.

Thanks,
Klas



^ permalink raw reply	[flat|nested] 17+ messages in thread

* Re: [Xenomai] Peak PCI CAN card RTCAN problem
  2012-06-28 16:09                   ` Klas Kronander
@ 2012-06-28 16:12                     ` Gilles Chanteperdrix
  2012-06-28 16:47                       ` Klas Kronander
  0 siblings, 1 reply; 17+ messages in thread
From: Gilles Chanteperdrix @ 2012-06-28 16:12 UTC (permalink / raw)
  To: Klas Kronander; +Cc: xenomai

On 06/28/2012 06:09 PM, Klas Kronander wrote:
> 
> 
> On 2012-06-28 15:27, Wolfgang Grandegger wrote:
>> On 06/28/2012 03:07 PM, Wolfgang Grandegger wrote:
>>> On 06/28/2012 02:52 PM, Klas Kronander wrote:
>>>> Hi,
>>>>
>>>> ok did that. The return value from rt_dev_socket is -14. I could not
>>>> find the definition of EPROTONOSUPPORT anywhere, so I don't know if this
>>>> was the expected value?
>>> ~$ errno 14
>>> #define	EFAULT		14	/* Bad address */
>>>
>>> Hm, strange. I have to look that up.
>> To better understand what you are doing... are you using a Xenomai
>> enabled kernel? Does the latency test work?
>>
>> Wolfgang.
> 
> Yes, I am using a Xenomai enabled kernel. None of the tests seem to work 
> though.
> 
> clocktest just gives me zeros. latency does not show anything. I have 
> tried to recompile the kernel, still the same result. How find out 
> what's wrong? I followed the guidlines on the xenomai website for the 
> kernel configuration.

http://www.xenomai.org/index.php/Configuring_x86_kernels#Does_Xenomai_work.3F

-- 
					    Gilles.


^ permalink raw reply	[flat|nested] 17+ messages in thread

* Re: [Xenomai] Peak PCI CAN card RTCAN problem
  2012-06-28 16:12                     ` Gilles Chanteperdrix
@ 2012-06-28 16:47                       ` Klas Kronander
  2012-06-28 16:48                         ` Gilles Chanteperdrix
  0 siblings, 1 reply; 17+ messages in thread
From: Klas Kronander @ 2012-06-28 16:47 UTC (permalink / raw)
  To: Gilles Chanteperdrix, xenomai

On 2012-06-28 18:12, Gilles Chanteperdrix wrote:
> On 06/28/2012 06:09 PM, Klas Kronander wrote:
>>
>> On 2012-06-28 15:27, Wolfgang Grandegger wrote:
>>> On 06/28/2012 03:07 PM, Wolfgang Grandegger wrote:
>>>> On 06/28/2012 02:52 PM, Klas Kronander wrote:
>>>>> Hi,
>>>>>
>>>>> ok did that. The return value from rt_dev_socket is -14. I could not
>>>>> find the definition of EPROTONOSUPPORT anywhere, so I don't know if this
>>>>> was the expected value?
>>>> ~$ errno 14
>>>> #define	EFAULT		14	/* Bad address */
>>>>
>>>> Hm, strange. I have to look that up.
>>> To better understand what you are doing... are you using a Xenomai
>>> enabled kernel? Does the latency test work?
>>>
>>> Wolfgang.
>> Yes, I am using a Xenomai enabled kernel. None of the tests seem to work
>> though.
>>
>> clocktest just gives me zeros. latency does not show anything. I have
>> tried to recompile the kernel, still the same result. How find out
>> what's wrong? I followed the guidlines on the xenomai website for the
>> kernel configuration.
> http://www.xenomai.org/index.php/Configuring_x86_kernels#Does_Xenomai_work.3F
>

In the syslog it seems as xenomai was initialized properly. The mesages 
correspond to those on the faq. I also looked through the 
troubleshooting file, nothing there seems to help my case. I will try to 
redo everything again with a different kernel version to see if that 
helps. Any other ideas?

Thanks,
Klas


^ permalink raw reply	[flat|nested] 17+ messages in thread

* Re: [Xenomai] Peak PCI CAN card RTCAN problem
  2012-06-28 16:47                       ` Klas Kronander
@ 2012-06-28 16:48                         ` Gilles Chanteperdrix
  2012-07-04 10:59                           ` Kronander Klas
  0 siblings, 1 reply; 17+ messages in thread
From: Gilles Chanteperdrix @ 2012-06-28 16:48 UTC (permalink / raw)
  To: Klas Kronander; +Cc: xenomai

On 06/28/2012 06:47 PM, Klas Kronander wrote:
> On 2012-06-28 18:12, Gilles Chanteperdrix wrote:
>> On 06/28/2012 06:09 PM, Klas Kronander wrote:
>>>
>>> On 2012-06-28 15:27, Wolfgang Grandegger wrote:
>>>> On 06/28/2012 03:07 PM, Wolfgang Grandegger wrote:
>>>>> On 06/28/2012 02:52 PM, Klas Kronander wrote:
>>>>>> Hi,
>>>>>>
>>>>>> ok did that. The return value from rt_dev_socket is -14. I could not
>>>>>> find the definition of EPROTONOSUPPORT anywhere, so I don't know if this
>>>>>> was the expected value?
>>>>> ~$ errno 14
>>>>> #define	EFAULT		14	/* Bad address */
>>>>>
>>>>> Hm, strange. I have to look that up.
>>>> To better understand what you are doing... are you using a Xenomai
>>>> enabled kernel? Does the latency test work?
>>>>
>>>> Wolfgang.
>>> Yes, I am using a Xenomai enabled kernel. None of the tests seem to work
>>> though.
>>>
>>> clocktest just gives me zeros. latency does not show anything. I have
>>> tried to recompile the kernel, still the same result. How find out
>>> what's wrong? I followed the guidlines on the xenomai website for the
>>> kernel configuration.
>> http://www.xenomai.org/index.php/Configuring_x86_kernels#Does_Xenomai_work.3F
>>
> 
> In the syslog it seems as xenomai was initialized properly. The mesages 
> correspond to those on the faq. I also looked through the 
> troubleshooting file, nothing there seems to help my case. I will try to 
> redo everything again with a different kernel version to see if that 
> helps. Any other ideas?

If you are using gcc 4.6.1, please pass CFLAGS=-fno-omit-frame-pointer
on the configure script command line.

-- 
					    Gilles.


^ permalink raw reply	[flat|nested] 17+ messages in thread

* Re: [Xenomai] Peak PCI CAN card RTCAN problem
  2012-06-28 16:48                         ` Gilles Chanteperdrix
@ 2012-07-04 10:59                           ` Kronander Klas
  2012-07-04 11:01                             ` Gilles Chanteperdrix
  0 siblings, 1 reply; 17+ messages in thread
From: Kronander Klas @ 2012-07-04 10:59 UTC (permalink / raw)
  To: Gilles Chanteperdrix; +Cc: xenomai

Hi,

I still have the same issues, with identical results with either of the four different kernel versions and corresponding adeos patches I have tried. 

After having booted into the kernel and installed the user-space xenomai libraries, none of the tests in /usr/xenomai/bin works. If i configure with --enable-dlopen-skins i get the following result from ./usr/xenomai/bin/clocktest:
== Tested clock: 0 (CLOCK_REALTIME)
CPU      ToD offset [us] ToD drift [us/s]      warps max delta [us]
--- -------------------- ---------------- ---------- --------------
  0                  0.0            0.000          0            0.0
  1                  0.0            0.000          0            0.0
  2                  0.0            0.000          0            0.0
  3                  0.0            0.000          0            0.0

If run the xenomai ./configure script without any arguments and then build and install, I get the following message when running ./usr/xenomai/bin/clocktest:
Xenomai Posix skin init: pthread_setschedparam: Function not implemented

As before, looking in the syslog it appears that Xenomai is up and running:
Jul  4 12:33:45 wam-desktop kernel: [    2.089906] I-pipe: Domain Xenomai registered.
Jul  4 12:33:45 wam-desktop kernel: [    2.089995] Xenomai: hal/i386 started.
Jul  4 12:33:45 wam-desktop kernel: [    2.090031] Xenomai: scheduling class idle registered.
Jul  4 12:33:45 wam-desktop kernel: [    2.090032] Xenomai: scheduling class rt registered.
Jul  4 12:33:45 wam-desktop kernel: [    2.090874] Xenomai: real-time nucleus v2.6.0 (Movin' On) loaded.
Jul  4 12:33:45 wam-desktop kernel: [    2.091059] Xenomai: starting native API services.
Jul  4 12:33:45 wam-desktop kernel: [    2.091061] Xenomai: starting POSIX services.
Jul  4 12:33:45 wam-desktop kernel: [    2.091085] Xenomai: starting RTDM services.


Thanks in advance for any ideas or suggestions on finding the issue,
Klas
________________________________________
From: Gilles Chanteperdrix [gilles.chanteperdrix@xenomai.org]
Sent: 28 June 2012 18:48
To: Kronander Klas
Cc: xenomai@xenomai.org
Subject: Re: [Xenomai] Peak PCI CAN card RTCAN problem

On 06/28/2012 06:47 PM, Klas Kronander wrote:
> On 2012-06-28 18:12, Gilles Chanteperdrix wrote:
>> On 06/28/2012 06:09 PM, Klas Kronander wrote:
>>>
>>> On 2012-06-28 15:27, Wolfgang Grandegger wrote:
>>>> On 06/28/2012 03:07 PM, Wolfgang Grandegger wrote:
>>>>> On 06/28/2012 02:52 PM, Klas Kronander wrote:
>>>>>> Hi,
>>>>>>
>>>>>> ok did that. The return value from rt_dev_socket is -14. I could not
>>>>>> find the definition of EPROTONOSUPPORT anywhere, so I don't know if this
>>>>>> was the expected value?
>>>>> ~$ errno 14
>>>>> #define   EFAULT          14      /* Bad address */
>>>>>
>>>>> Hm, strange. I have to look that up.
>>>> To better understand what you are doing... are you using a Xenomai
>>>> enabled kernel? Does the latency test work?
>>>>
>>>> Wolfgang.
>>> Yes, I am using a Xenomai enabled kernel. None of the tests seem to work
>>> though.
>>>
>>> clocktest just gives me zeros. latency does not show anything. I have
>>> tried to recompile the kernel, still the same result. How find out
>>> what's wrong? I followed the guidlines on the xenomai website for the
>>> kernel configuration.
>> http://www.xenomai.org/index.php/Configuring_x86_kernels#Does_Xenomai_work.3F
>>
>
> In the syslog it seems as xenomai was initialized properly. The mesages
> correspond to those on the faq. I also looked through the
> troubleshooting file, nothing there seems to help my case. I will try to
> redo everything again with a different kernel version to see if that
> helps. Any other ideas?

If you are using gcc 4.6.1, please pass CFLAGS=-fno-omit-frame-pointer
on the configure script command line.

--
                                            Gilles.


^ permalink raw reply	[flat|nested] 17+ messages in thread

* Re: [Xenomai] Peak PCI CAN card RTCAN problem
  2012-07-04 10:59                           ` Kronander Klas
@ 2012-07-04 11:01                             ` Gilles Chanteperdrix
  0 siblings, 0 replies; 17+ messages in thread
From: Gilles Chanteperdrix @ 2012-07-04 11:01 UTC (permalink / raw)
  To: Kronander Klas; +Cc: xenomai

On 07/04/2012 12:59 PM, Kronander Klas wrote:
> If run the xenomai ./configure script without any arguments and then
> build and install, I get the following message when running
> ./usr/xenomai/bin/clocktest:

Have you tried to run ./configure CFLAGS="-fno-omit-frame-pointer" as I
already suggested?

-- 
                                                                Gilles.


^ permalink raw reply	[flat|nested] 17+ messages in thread

end of thread, other threads:[~2012-07-04 11:01 UTC | newest]

Thread overview: 17+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2012-06-23 14:24 [Xenomai] Peak PCI CAN card RTCAN problem Kronander Klas
2012-06-23 17:49 ` Gilles Chanteperdrix
2012-06-28 11:11   ` Kronander Klas
2012-06-28 11:28     ` Wolfgang Grandegger
2012-06-28 11:38       ` Wolfgang Grandegger
2012-06-28 11:45         ` Kronander Klas
2012-06-28 11:57           ` Wolfgang Grandegger
2012-06-28 12:52             ` Klas Kronander
2012-06-28 13:07               ` Wolfgang Grandegger
2012-06-28 13:27                 ` Wolfgang Grandegger
2012-06-28 16:09                   ` Klas Kronander
2012-06-28 16:12                     ` Gilles Chanteperdrix
2012-06-28 16:47                       ` Klas Kronander
2012-06-28 16:48                         ` Gilles Chanteperdrix
2012-07-04 10:59                           ` Kronander Klas
2012-07-04 11:01                             ` Gilles Chanteperdrix
2012-06-28 12:45     ` Gilles Chanteperdrix

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.