All of lore.kernel.org
 help / color / mirror / Atom feed
* [Xenomai-core] zombie thread would not die
@ 2009-03-01 17:30 Niklaus Giger
  2009-03-01 17:39 ` Philippe Gerum
  0 siblings, 1 reply; 2+ messages in thread
From: Niklaus Giger @ 2009-03-01 17:30 UTC (permalink / raw)
  To: xenomai-core

Hi

I successfully ported our basic IO library to Xenomai with the vxWorks
skin running on our PPC405GPr based HCU4 (which was accepted into Linux 
2.6.28). I just added support for one external interrrupt (occurring every 
millisecond) and two CAN interrupts (not yet currently activated).

Now my program starts up, recognizes the HW but stops sometimes after 5 
seconds, sometimes after 30 seconds. Always with a zombie message. Can anyboy 
tell me, where the error is or what I have to check?

I am using  svn://svn.gna.org/svn/xenomai/trunk (Revision 4662).
uname -a gives
Linux hcu4 2.6.28.5-dirty #12 Sun Mar 1 13:45:44 CET 2009 ppc ppc ppc 
GNU/Linux
cat /proc/ipipe/version
2.4-04
cat /proc/interrupts
           CPU0
 16:        220   UIC   Level     serial
 17:       7971   UIC   Level     MAL TX EOB
 18:      12537   UIC   Level     MAL RX EOB
 19:          0   UIC   Level     CAN1
 20:          0   UIC   Level     CAN2
 21:          0   UIC   Level     MAL SERR
 22:          0   UIC   Level     MAL TX DE
 23:          0   UIC   Level     MAL RX DE
 24:          0   UIC   Level     EMAC
 25:          0   UIC   Level     CPLD
BAD:          0


 Xenomai: fatal: zombie thread iTsk_25 (c18e3e20) would not die...
 CPU  PID    PRI      TIMEOUT  STAT      NAME
>  0  0      231      0        00200080  ROOT
   0  1032   211(45)  1        00100184  root
   0  1034   231(25)  0        00100380  tUtilLog
   0  1036   246(10)  0        00100182  tBBprio
   0  1039   216(40)  0        00100182  BSysThreadPool0
   0  1040   216(40)  0        00100182  BSysThreadPool1
   0  1041   216(40)  0        00100182  BSysThreadPool2
   0  1042   216(40)  0        00100182  BSysThreadPool3
   0  1043   216(40)  0        00100182  BSysThreadPool4
   0  1044   216(40)  0        00100182  BSysThreadPool5
   0  1045   216(40)  0        00100182  BSysThreadPool6
   0  1046   216(40)  0        00100182  BSysThreadPool7
   0  1047   216(40)  0        00100182  BSysThreadPool8
   0  1048   216(40)  0        00100182  BSysThreadPool9
   0  1049   254(2)   2        00100186  tBBmsec
Master time base: clock=273984265688

Call Trace:
[c1a8de50] [c0006c7c] show_stack+0x44/0x16c (unreliable)
[c1a8de90] [c00584d8] __xnpod_schedule+0x444/0x634
[c1a8dee0] [c00546b8] xnintr_clock_handler+0x1c4/0x1d8
[c1a8df10] [c004f844] __ipipe_dispatch_wired_nocheck+0x80/0xf4
[c1a8df30] [c0008a74] __ipipe_grab_timer+0x78/0xd4
[c1a8df40] [c0010244] __ipipe_ret_from_except+0x0/0xc

Best regards

-- 
Niklaus Giger



^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: [Xenomai-core] zombie thread would not die
  2009-03-01 17:30 [Xenomai-core] zombie thread would not die Niklaus Giger
@ 2009-03-01 17:39 ` Philippe Gerum
  0 siblings, 0 replies; 2+ messages in thread
From: Philippe Gerum @ 2009-03-01 17:39 UTC (permalink / raw)
  To: Niklaus Giger; +Cc: xenomai-core

Niklaus Giger wrote:
> Hi
> 
> I successfully ported our basic IO library to Xenomai with the vxWorks
> skin running on our PPC405GPr based HCU4 (which was accepted into Linux 
> 2.6.28). I just added support for one external interrrupt (occurring every 
> millisecond) and two CAN interrupts (not yet currently activated).
> 
> Now my program starts up, recognizes the HW but stops sometimes after 5 
> seconds, sometimes after 30 seconds. Always with a zombie message. Can anyboy 
> tell me, where the error is or what I have to check?
> 
> I am using  svn://svn.gna.org/svn/xenomai/trunk (Revision 4662).
> uname -a gives
> Linux hcu4 2.6.28.5-dirty #12 Sun Mar 1 13:45:44 CET 2009 ppc ppc ppc 
> GNU/Linux
> cat /proc/ipipe/version
> 2.4-04
> cat /proc/interrupts

Could you check with Xenomai 2.4.7 as well?

>            CPU0
>  16:        220   UIC   Level     serial
>  17:       7971   UIC   Level     MAL TX EOB
>  18:      12537   UIC   Level     MAL RX EOB
>  19:          0   UIC   Level     CAN1
>  20:          0   UIC   Level     CAN2
>  21:          0   UIC   Level     MAL SERR
>  22:          0   UIC   Level     MAL TX DE
>  23:          0   UIC   Level     MAL RX DE
>  24:          0   UIC   Level     EMAC
>  25:          0   UIC   Level     CPLD
> BAD:          0
> 
> 
>  Xenomai: fatal: zombie thread iTsk_25 (c18e3e20) would not die...

Is this task expected to exit during the test, or to be deleted?

>  CPU  PID    PRI      TIMEOUT  STAT      NAME
>>  0  0      231      0        00200080  ROOT
>    0  1032   211(45)  1        00100184  root
>    0  1034   231(25)  0        00100380  tUtilLog
>    0  1036   246(10)  0        00100182  tBBprio
>    0  1039   216(40)  0        00100182  BSysThreadPool0
>    0  1040   216(40)  0        00100182  BSysThreadPool1
>    0  1041   216(40)  0        00100182  BSysThreadPool2
>    0  1042   216(40)  0        00100182  BSysThreadPool3
>    0  1043   216(40)  0        00100182  BSysThreadPool4
>    0  1044   216(40)  0        00100182  BSysThreadPool5
>    0  1045   216(40)  0        00100182  BSysThreadPool6
>    0  1046   216(40)  0        00100182  BSysThreadPool7
>    0  1047   216(40)  0        00100182  BSysThreadPool8
>    0  1048   216(40)  0        00100182  BSysThreadPool9
>    0  1049   254(2)   2        00100186  tBBmsec
> Master time base: clock=273984265688
> 
> Call Trace:
> [c1a8de50] [c0006c7c] show_stack+0x44/0x16c (unreliable)
> [c1a8de90] [c00584d8] __xnpod_schedule+0x444/0x634
> [c1a8dee0] [c00546b8] xnintr_clock_handler+0x1c4/0x1d8
> [c1a8df10] [c004f844] __ipipe_dispatch_wired_nocheck+0x80/0xf4
> [c1a8df30] [c0008a74] __ipipe_grab_timer+0x78/0xd4
> [c1a8df40] [c0010244] __ipipe_ret_from_except+0x0/0xc
> 
> Best regards
> 


-- 
Philippe.


^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2009-03-01 17:39 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2009-03-01 17:30 [Xenomai-core] zombie thread would not die Niklaus Giger
2009-03-01 17:39 ` Philippe Gerum

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.