All of lore.kernel.org
 help / color / mirror / Atom feed
* [Xenomai-help] Meaning of cache locking in /proc/xenomai/faults
@ 2009-03-27  9:00 Axel Beierlein
  2009-03-27  9:25 ` Philippe Gerum
  0 siblings, 1 reply; 6+ messages in thread
From: Axel Beierlein @ 2009-03-27  9:00 UTC (permalink / raw)
  To: xenomai

Hello,

on my MPC5200 i have the following output of /proc/xenomai/faults

TRAP         CPU0
   0:            0    (Data or instruction access)
   1:            0    (Alignment)
   2:            0    (Altivec unavailable)
   3:            0    (Program check exception)
   4:            0    (Machine check exception)
   5:            0    (Unknown)
   6:            0    (Instruction breakpoint)
   7:            0    (Run mode exception)
   8:            0    (Single-step exception)
   9:            0    (Non-recoverable exception)
  10:            0    (Software emulation)
  11:            0    (Debug)
  12:            0    (SPE)
  13:            0    (Altivec assist)
  14:  -1072753084    (Cache-locking exception)
  15:            0    (Kernel FP unavailable)

Now i am trying to find out what the negative Value of Cache-locking and 
what this Trap general mean and when it was triggered.

Can you give me some advice?

TIA


^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [Xenomai-help] Meaning of cache locking in /proc/xenomai/faults
  2009-03-27  9:00 [Xenomai-help] Meaning of cache locking in /proc/xenomai/faults Axel Beierlein
@ 2009-03-27  9:25 ` Philippe Gerum
       [not found]   ` <49CC9E46.3050501@domain.hid>
  0 siblings, 1 reply; 6+ messages in thread
From: Philippe Gerum @ 2009-03-27  9:25 UTC (permalink / raw)
  To: belatronix; +Cc: xenomai

On Fri, 2009-03-27 at 10:00 +0100, Axel Beierlein wrote:
> Hello,
> 
> on my MPC5200 i have the following output of /proc/xenomai/faults
> 
> TRAP         CPU0
>    0:            0    (Data or instruction access)
>    1:            0    (Alignment)
>    2:            0    (Altivec unavailable)
>    3:            0    (Program check exception)
>    4:            0    (Machine check exception)
>    5:            0    (Unknown)
>    6:            0    (Instruction breakpoint)
>    7:            0    (Run mode exception)
>    8:            0    (Single-step exception)
>    9:            0    (Non-recoverable exception)
>   10:            0    (Software emulation)
>   11:            0    (Debug)
>   12:            0    (SPE)
>   13:            0    (Altivec assist)
>   14:  -1072753084    (Cache-locking exception)
>   15:            0    (Kernel FP unavailable)
> 
> Now i am trying to find out what the negative Value of Cache-locking and 
> what this Trap general mean and when it was triggered.

The negative value is only the sign of an overflow. However, a cache
locking exception is specific to FSL_BOOKE archs. What is your kernel
release and exact ppc platform configuration?

> 
> Can you give me some advice?
> 
> TIA
> 
> _______________________________________________
> Xenomai-help mailing list
> Xenomai-help@domain.hid
> https://mail.gna.org/listinfo/xenomai-help
-- 
Philippe.




^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [Xenomai-help] Meaning of cache locking in /proc/xenomai/faults
       [not found]   ` <49CC9E46.3050501@domain.hid>
@ 2009-03-27 10:28     ` Philippe Gerum
  2009-03-27 13:11       ` Axel Beierlein
  0 siblings, 1 reply; 6+ messages in thread
From: Philippe Gerum @ 2009-03-27 10:28 UTC (permalink / raw)
  To: belatronix; +Cc: xenomai

On Fri, 2009-03-27 at 10:37 +0100, Axel Beierlein wrote:
> Philippe Gerum schrieb:
> > On Fri, 2009-03-27 at 10:00 +0100, Axel Beierlein wrote:
> >> Hello,
> >>
> >> on my MPC5200 i have the following output of /proc/xenomai/faults
> >>
> >> TRAP         CPU0
> >>    0:            0    (Data or instruction access)
> >>    1:            0    (Alignment)
> >>    2:            0    (Altivec unavailable)
> >>    3:            0    (Program check exception)
> >>    4:            0    (Machine check exception)
> >>    5:            0    (Unknown)
> >>    6:            0    (Instruction breakpoint)
> >>    7:            0    (Run mode exception)
> >>    8:            0    (Single-step exception)
> >>    9:            0    (Non-recoverable exception)
> >>   10:            0    (Software emulation)
> >>   11:            0    (Debug)
> >>   12:            0    (SPE)
> >>   13:            0    (Altivec assist)
> >>   14:  -1072753084    (Cache-locking exception)
> >>   15:            0    (Kernel FP unavailable)
> >>
> >> Now i am trying to find out what the negative Value of Cache-locking and 
> >> what this Trap general mean and when it was triggered.
> > 
> > The negative value is only the sign of an overflow. However, a cache
> > locking exception is specific to FSL_BOOKE archs. What is your kernel
> > release and exact ppc platform configuration?


You can ignore the last two counters, when 2.4.25-ppc is involved, they
contain garbage in fact; this does not have any consequences beyond that
strange output. The patch below should fix them, actually leave them
zeroed.

--- include/asm-ppc/ipipe.h~	2008-06-08 16:15:51.000000000 +0200
+++ include/asm-ppc/ipipe.h	2009-03-27 11:22:31.000000000 +0100
@@ -80,7 +80,7 @@ do {						\
 #define IPIPE_TRAP_DEBUG	11	/* Debug exception */
 #define IPIPE_TRAP_SPE		12	/* SPE exception */
 #define IPIPE_TRAP_ALTASSIST	13	/* Altivec assist exception */
-#define IPIPE_NR_FAULTS		14
+#define IPIPE_NR_FAULTS		16
 /* Pseudo-vectors used for kernel events */
 #define IPIPE_FIRST_EVENT	IPIPE_NR_FAULTS
 #define IPIPE_EVENT_SYSCALL	(IPIPE_FIRST_EVENT)

-- 
Philippe.




^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [Xenomai-help] Meaning of cache locking in /proc/xenomai/faults
  2009-03-27 10:28     ` Philippe Gerum
@ 2009-03-27 13:11       ` Axel Beierlein
  2009-03-27 13:46         ` Philippe Gerum
  0 siblings, 1 reply; 6+ messages in thread
From: Axel Beierlein @ 2009-03-27 13:11 UTC (permalink / raw)
  To: Philippe Gerum, xenomai

Philippe Gerum schrieb:
> On Fri, 2009-03-27 at 10:37 +0100, Axel Beierlein wrote:
>> Philippe Gerum schrieb:
>>> On Fri, 2009-03-27 at 10:00 +0100, Axel Beierlein wrote:
>>>> Hello,
>>>>
>>>> on my MPC5200 i have the following output of /proc/xenomai/faults
>>>>
>>>> TRAP         CPU0
>>>>    0:            0    (Data or instruction access)
>>>>    1:            0    (Alignment)
>>>>    2:            0    (Altivec unavailable)
>>>>    3:            0    (Program check exception)
>>>>    4:            0    (Machine check exception)
>>>>    5:            0    (Unknown)
>>>>    6:            0    (Instruction breakpoint)
>>>>    7:            0    (Run mode exception)
>>>>    8:            0    (Single-step exception)
>>>>    9:            0    (Non-recoverable exception)
>>>>   10:            0    (Software emulation)
>>>>   11:            0    (Debug)
>>>>   12:            0    (SPE)
>>>>   13:            0    (Altivec assist)
>>>>   14:  -1072753084    (Cache-locking exception)
>>>>   15:            0    (Kernel FP unavailable)
>>>>
>>>> Now i am trying to find out what the negative Value of Cache-locking and 
>>>> what this Trap general mean and when it was triggered.
>>> The negative value is only the sign of an overflow. However, a cache
>>> locking exception is specific to FSL_BOOKE archs. What is your kernel
>>> release and exact ppc platform configuration?
> 
> 
> You can ignore the last two counters, when 2.4.25-ppc is involved, they
> contain garbage in fact; this does not have any consequences beyond that
> strange output. The patch below should fix them, actually leave them
> zeroed.
> 
> --- include/asm-ppc/ipipe.h~	2008-06-08 16:15:51.000000000 +0200
> +++ include/asm-ppc/ipipe.h	2009-03-27 11:22:31.000000000 +0100
> @@ -80,7 +80,7 @@ do {						\
>  #define IPIPE_TRAP_DEBUG	11	/* Debug exception */
>  #define IPIPE_TRAP_SPE		12	/* SPE exception */
>  #define IPIPE_TRAP_ALTASSIST	13	/* Altivec assist exception */
> -#define IPIPE_NR_FAULTS		14
> +#define IPIPE_NR_FAULTS		16
>  /* Pseudo-vectors used for kernel events */
>  #define IPIPE_FIRST_EVENT	IPIPE_NR_FAULTS
>  #define IPIPE_EVENT_SYSCALL	(IPIPE_FIRST_EVENT)
> 

Thank you Philippe for your Help.

By the way, can you, or someone else, point me to a message where are 
the other traps are explained or can you give me short examples for 
situations where the listed exceptions are triggered?
For example: What is an alignment Trap and which situation can trigger 
such a fault.

Axel


^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [Xenomai-help] Meaning of cache locking in /proc/xenomai/faults
  2009-03-27 13:11       ` Axel Beierlein
@ 2009-03-27 13:46         ` Philippe Gerum
  2009-03-27 14:15           ` Axel Beierlein
  0 siblings, 1 reply; 6+ messages in thread
From: Philippe Gerum @ 2009-03-27 13:46 UTC (permalink / raw)
  To: belatronix; +Cc: xenomai

On Fri, 2009-03-27 at 14:11 +0100, Axel Beierlein wrote:
> Philippe Gerum schrieb:
> > On Fri, 2009-03-27 at 10:37 +0100, Axel Beierlein wrote:
> >> Philippe Gerum schrieb:
> >>> On Fri, 2009-03-27 at 10:00 +0100, Axel Beierlein wrote:
> >>>> Hello,
> >>>>
> >>>> on my MPC5200 i have the following output of /proc/xenomai/faults
> >>>>
> >>>> TRAP         CPU0
> >>>>    0:            0    (Data or instruction access)
> >>>>    1:            0    (Alignment)
> >>>>    2:            0    (Altivec unavailable)
> >>>>    3:            0    (Program check exception)
> >>>>    4:            0    (Machine check exception)
> >>>>    5:            0    (Unknown)
> >>>>    6:            0    (Instruction breakpoint)
> >>>>    7:            0    (Run mode exception)
> >>>>    8:            0    (Single-step exception)
> >>>>    9:            0    (Non-recoverable exception)
> >>>>   10:            0    (Software emulation)
> >>>>   11:            0    (Debug)
> >>>>   12:            0    (SPE)
> >>>>   13:            0    (Altivec assist)
> >>>>   14:  -1072753084    (Cache-locking exception)
> >>>>   15:            0    (Kernel FP unavailable)
> >>>>
> >>>> Now i am trying to find out what the negative Value of Cache-locking and 
> >>>> what this Trap general mean and when it was triggered.
> >>> The negative value is only the sign of an overflow. However, a cache
> >>> locking exception is specific to FSL_BOOKE archs. What is your kernel
> >>> release and exact ppc platform configuration?
> > 
> > 
> > You can ignore the last two counters, when 2.4.25-ppc is involved, they
> > contain garbage in fact; this does not have any consequences beyond that
> > strange output. The patch below should fix them, actually leave them
> > zeroed.
> > 
> > --- include/asm-ppc/ipipe.h~	2008-06-08 16:15:51.000000000 +0200
> > +++ include/asm-ppc/ipipe.h	2009-03-27 11:22:31.000000000 +0100
> > @@ -80,7 +80,7 @@ do {						\
> >  #define IPIPE_TRAP_DEBUG	11	/* Debug exception */
> >  #define IPIPE_TRAP_SPE		12	/* SPE exception */
> >  #define IPIPE_TRAP_ALTASSIST	13	/* Altivec assist exception */
> > -#define IPIPE_NR_FAULTS		14
> > +#define IPIPE_NR_FAULTS		16
> >  /* Pseudo-vectors used for kernel events */
> >  #define IPIPE_FIRST_EVENT	IPIPE_NR_FAULTS
> >  #define IPIPE_EVENT_SYSCALL	(IPIPE_FIRST_EVENT)
> > 
> 
> Thank you Philippe for your Help.
> 
> By the way, can you, or someone else, point me to a message where are 
> the other traps are explained or can you give me short examples for 
> situations where the listed exceptions are triggered?
> For example: What is an alignment Trap and which situation can trigger 
> such a fault.
> 

You probably want to have alook at the PowerPC UISA documentation:
http://download.boulder.ibm.com/ibmdl/pub/software/dw/library/es-ppcbook1.zip

You will find common reasons for traps. Additionally, you should refer
to your core's reference manual, since it may have specifics regarding
why/when some exceptions are triggered. Alignment exception is usually
due to data placement, misaligned float ops are usual suspects, but
depending on your particular ppc core, other reasons for alignment traps
may exist (e.g. e300 cores fire this exception when using a particular
byte-zeroing instruction on cache-inhibited memory). In short, YMMV.

> Axel
-- 
Philippe.




^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [Xenomai-help] Meaning of cache locking in /proc/xenomai/faults
  2009-03-27 13:46         ` Philippe Gerum
@ 2009-03-27 14:15           ` Axel Beierlein
  0 siblings, 0 replies; 6+ messages in thread
From: Axel Beierlein @ 2009-03-27 14:15 UTC (permalink / raw)
  To: Philippe Gerum, xenomai

Philippe Gerum schrieb:
> On Fri, 2009-03-27 at 14:11 +0100, Axel Beierlein wrote:
>> Philippe Gerum schrieb:
>>> On Fri, 2009-03-27 at 10:37 +0100, Axel Beierlein wrote:
>>>> Philippe Gerum schrieb:
>>>>> On Fri, 2009-03-27 at 10:00 +0100, Axel Beierlein wrote:
>>>>>> Hello,
>>>>>>
>>>>>> on my MPC5200 i have the following output of /proc/xenomai/faults
>>>>>>
>>>>>> TRAP         CPU0
>>>>>>    0:            0    (Data or instruction access)
>>>>>>    1:            0    (Alignment)
>>>>>>    2:            0    (Altivec unavailable)
>>>>>>    3:            0    (Program check exception)
>>>>>>    4:            0    (Machine check exception)
>>>>>>    5:            0    (Unknown)
>>>>>>    6:            0    (Instruction breakpoint)
>>>>>>    7:            0    (Run mode exception)
>>>>>>    8:            0    (Single-step exception)
>>>>>>    9:            0    (Non-recoverable exception)
>>>>>>   10:            0    (Software emulation)
>>>>>>   11:            0    (Debug)
>>>>>>   12:            0    (SPE)
>>>>>>   13:            0    (Altivec assist)
>>>>>>   14:  -1072753084    (Cache-locking exception)
>>>>>>   15:            0    (Kernel FP unavailable)
>>>>>>
>>>>>> Now i am trying to find out what the negative Value of Cache-locking and 
>>>>>> what this Trap general mean and when it was triggered.
>>>>> The negative value is only the sign of an overflow. However, a cache
>>>>> locking exception is specific to FSL_BOOKE archs. What is your kernel
>>>>> release and exact ppc platform configuration?
>>>
>>> You can ignore the last two counters, when 2.4.25-ppc is involved, they
>>> contain garbage in fact; this does not have any consequences beyond that
>>> strange output. The patch below should fix them, actually leave them
>>> zeroed.
>>>
>>> --- include/asm-ppc/ipipe.h~	2008-06-08 16:15:51.000000000 +0200
>>> +++ include/asm-ppc/ipipe.h	2009-03-27 11:22:31.000000000 +0100
>>> @@ -80,7 +80,7 @@ do {						\
>>>  #define IPIPE_TRAP_DEBUG	11	/* Debug exception */
>>>  #define IPIPE_TRAP_SPE		12	/* SPE exception */
>>>  #define IPIPE_TRAP_ALTASSIST	13	/* Altivec assist exception */
>>> -#define IPIPE_NR_FAULTS		14
>>> +#define IPIPE_NR_FAULTS		16
>>>  /* Pseudo-vectors used for kernel events */
>>>  #define IPIPE_FIRST_EVENT	IPIPE_NR_FAULTS
>>>  #define IPIPE_EVENT_SYSCALL	(IPIPE_FIRST_EVENT)
>>>
>> Thank you Philippe for your Help.
>>
>> By the way, can you, or someone else, point me to a message where are 
>> the other traps are explained or can you give me short examples for 
>> situations where the listed exceptions are triggered?
>> For example: What is an alignment Trap and which situation can trigger 
>> such a fault.
>>
> 
> You probably want to have alook at the PowerPC UISA documentation:
> http://download.boulder.ibm.com/ibmdl/pub/software/dw/library/es-ppcbook1.zip
> 
> You will find common reasons for traps. Additionally, you should refer
> to your core's reference manual, since it may have specifics regarding
> why/when some exceptions are triggered. Alignment exception is usually
> due to data placement, misaligned float ops are usual suspects, but
> depending on your particular ppc core, other reasons for alignment traps
> may exist (e.g. e300 cores fire this exception when using a particular
> byte-zeroing instruction on cache-inhibited memory). In short, YMMV.
> 
>> Axel

Yeah! Thats it. Thanks. I find all the needed informations in the 
e300coreRM Manual from Freescale.

Axel



^ permalink raw reply	[flat|nested] 6+ messages in thread

end of thread, other threads:[~2009-03-27 14:15 UTC | newest]

Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2009-03-27  9:00 [Xenomai-help] Meaning of cache locking in /proc/xenomai/faults Axel Beierlein
2009-03-27  9:25 ` Philippe Gerum
     [not found]   ` <49CC9E46.3050501@domain.hid>
2009-03-27 10:28     ` Philippe Gerum
2009-03-27 13:11       ` Axel Beierlein
2009-03-27 13:46         ` Philippe Gerum
2009-03-27 14:15           ` Axel Beierlein

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.