[asterisk-users] TE410P (1st) without cables always green

Marcio Gomes mpglista at microlink.com.br
Thu Feb 9 10:11:12 CST 2012


Hello Shaun,

1) dahdi-linux-complete-2.2.0.2+2.2.0 , is generatiing 1K int/s after 
patches..

2) Looking in 2.2.1 code, i see to wct4xxp.c the alarmdebounce set in 
2500 and in 2.2.0 the alarmdebounce seted to 0,

I load the module with alarmdebouce=0 paramter => the interrupts are up 
to 1000 int/s

After this, I test all versions until dahdi-linux-complete-2.6.0+2.6.0

cd version
make distclean
make
make install
/etc/init.d/dahdi stop ;sleep 3; modprobe dahdi;  modprobe wct4xxp 
debug=1 alarmdebounce=0  ; dahdi_cfg
.itop =>  to see interrupts.

always are 1000 ints/s

I not touch in  losalarmdebounce=0 aisalarmdebounce=0 yelalarmdebounce=0 
module paramters.


3) Can you explain-me aboute how the alarmdebounce and its variants ( 
LOS, YEL, etcc ) impacts the system perfomance and how
can i fine tune this ?


Regards,

Marcio Gomes


Em 09/02/2012 10:45, Marcio Gomes escreveu:
> Hello Shaun,
>
> I take this approach :
>
> - zaptel 1.2.XX , 1.4.XX  I can see interrupts working  , I made some 
> patches to compile in 2.6.32 kernel tree.
>
> - dahdi 2.0.0 to dahdi-2.2.0, I can see interrupts working, I made 
> some patches to compile in 2.6.32 kernel tree
>
> - This "itop" from 2.2.0 tree :
>
>                                                                 
> IRQs/Second
>               Device (IRQ)             CPU0            CPU1           
> TOTAL
>                 rtc0 (  8):               0               
> 0               0
>                i8042 (  1):               0               
> 0               0
>                 eth0 ( 28):               3               
> 3               6
>                timer (  0):             435             566            
> 1001
>   ata_piix, ata_piix ( 19):               2               
> 1               3
>         pata_jmicron ( 16):               0               
> 0               0
>              cascade (  2):               0               
> 0               0
>              wct4xxp ( 20):             566             435            
> 1001
>
> As you can see, I am having 1K int/s
>
>
> - My patches are related to add #include <linux/sched.c> in 
> dahdi-base.c , wct4xxp.c and wctdm.c , and comment out the lines 
> relative to others drivers ( I only needs wcfxo, wctdm and wct4xxp
> to this box )
>
> - 2.2.1 and greater -> Compiles without patches, but not working, 
> int/s = 0
>
> I do not knows the main changes from 2.2.0 to to 2.2.1 dahdi-tree , 
> but it is clear to me that problems are in this changes. I can goes up 
> and down between versions
> with same issues, I do this because initially suspected  from da dead 
> lock in hardware create for some tests I have do.
>
>
> Regards,
>
> Marcio Gomes
>
>
> Em 07/02/2012 17:38, Marcio Gomes escreveu:
>> News from Zaptel..
>>
>> I patch the last zaptel.1.2.27 to compiles in 2.6.33 kernels 
>> series.... Surprise !!!!!!
>>
>>  20:      68011      71871   IO-APIC-fasteoi   wct4xxp
>>
>> in zttool the spans are RED, all is working with pre 1.4 zaptel and 
>> dahdi.
>>
>> The problems are in changes between zaptel.1.2.X and zaptel.1.4.X.
>>
>> I thinking now that is an old issue added from 1.2 to 1.4 branches 
>> and sucessivily to dahdi branches,
>> and I lost something in past.
>>
>> But I want go to asterisk 1.8 and greater with new drivers , Can 
>> anyone help me in this ?
>>
>> regards,
>>
>>
>>
>>
>> Em 07/02/2012 17:05, Marcio Gomes escreveu:
>>> Hello All,
>>>
>>>
>>>> I really not understand de APIC changes in IO-APIC-XXXX  and the 
>>>> greater 100 Interrupt numbers in old setups.. can you help me ?
>>>
>>>
>>> I forget to say,
>>>
>>> In production BOX, I "only" do a linux -R newconfig && reboot ,
>>>
>>> slots , hardware, setup, etc.. is the same.. no changes. It's give 
>>> me a confirmation about it is not a hardware issue, the only changes
>>> are in software.
>>>
>>> In the development box, a hardware problem iss possible, I will make 
>>> a copy of the working software and put in it to clarify this
>>> doubt.
>>>
>>> Regards,
>>>
>>> Marcio Gomes
>>>
>>>
>>>
>>> -- 
>>> _____________________________________________________________________
>>> -- Bandwidth and Colocation Provided by http://www.api-digital.com --
>>> New to Asterisk? Join us for a live introductory webinar every Thurs:
>>>               http://www.asterisk.org/hello
>>>
>>> asterisk-users mailing list
>>> To UNSUBSCRIBE or update options visit:
>>>   http://lists.digium.com/mailman/listinfo/asterisk-users
>>
>>
>> -- 
>> _____________________________________________________________________
>> -- Bandwidth and Colocation Provided by http://www.api-digital.com --
>> New to Asterisk? Join us for a live introductory webinar every Thurs:
>>               http://www.asterisk.org/hello
>>
>> asterisk-users mailing list
>> To UNSUBSCRIBE or update options visit:
>>   http://lists.digium.com/mailman/listinfo/asterisk-users
>
>
> -- 
> _____________________________________________________________________
> -- Bandwidth and Colocation Provided by http://www.api-digital.com --
> New to Asterisk? Join us for a live introductory webinar every Thurs:
>               http://www.asterisk.org/hello
>
> asterisk-users mailing list
> To UNSUBSCRIBE or update options visit:
>   http://lists.digium.com/mailman/listinfo/asterisk-users




More information about the asterisk-users mailing list