SCI interrupt handler : spend too much time

SCI interrupt handler : spend too much time

[roychris]

Hi,

Since 2.6, when I read /proc/acpi/thermal/THR*/temperature and battery
info, my system spends too much time in SCI interrupt handler and it
loses the sync of my mouse, a synaptic. Any issue ?

thx




-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive?  Does it
help you create better code?  SHARE THE LOVE, and help us help
YOU!  Click Here: http://sourceforge.net/donate/

Re: SCI interrupt handler : spend too much time

[Pavel Machek]

Hi!

> Since 2.6, when I read /proc/acpi/thermal/THR*/temperature and battery
> info, my system spends too much time in SCI interrupt handler and it
> loses the sync of my mouse, a synaptic. Any issue ?

Can you repeat it? Losing sync of mouse is probably a bug in inputs,
and should be fixed.
								Pavel
-- 
When do you have a heart between your knees?
[Johanka's followup: and *two* hearts?]


-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive?  Does it
help you create better code?  SHARE THE LOVE, and help us help
YOU!  Click Here: http://sourceforge.net/donate/

RE: SCI interrupt handler : spend too much time

[Yu, Luming]

It sound like serious one, if SCI interrupt handler was blamed for this issue.     --Luming

-----Original Message-----
From: acpi-devel-admin-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org [mailto:acpi-devel-admin-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org]On Behalf Of Pavel Machek
Sent: 2003?12?2? 2:28
To: roychris
Cc: acpi-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
Subject: Re: [ACPI] SCI interrupt handler : spend too much time


Hi!

> Since 2.6, when I read /proc/acpi/thermal/THR*/temperature and battery
> info, my system spends too much time in SCI interrupt handler and it
> loses the sync of my mouse, a synaptic. Any issue ?

Can you repeat it? Losing sync of mouse is probably a bug in inputs,
and should be fixed.
								Pavel
-- 
When do you have a heart between your knees?
[Johanka's followup: and *two* hearts?]


-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive?  Does it
help you create better code?  SHARE THE LOVE, and help us help
YOU!  Click Here: http://sourceforge.net/donate/
_______________________________________________
Acpi-devel mailing list
Acpi-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
https://lists.sourceforge.net/lists/listinfo/acpi-devel


-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive?  Does it
help you create better code?  SHARE THE LOVE, and help us help
YOU!  Click Here: http://sourceforge.net/donate/

Re: SCI interrupt handler : spend too much time

[roychris]

[-- Attachment #1: Type: text/plain, Size: 1063 bytes --]

I have contacted the developper of synaptic driver, and for sure it's 
the fault of SCI interrupt handler.
I can reproduce it because I never lose sync when I don't access to 
/proc/.../temperature and when I access it, exactly at the same time, I 
lose the sync.

Yu, Luming wrote:

>It sound like serious one, if SCI interrupt handler was blamed for this issue.     --Luming
>
>-----Original Message-----
>From: acpi-devel-admin-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org [mailto:acpi-devel-admin@lists.sourceforge.net]On Behalf Of Pavel Machek
>Sent: 2003?12?2? 2:28
>To: roychris
>Cc: acpi-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
>Subject: Re: [ACPI] SCI interrupt handler : spend too much time
>
>
>Hi!
>
>  
>
>>Since 2.6, when I read /proc/acpi/thermal/THR*/temperature and battery
>>info, my system spends too much time in SCI interrupt handler and it
>>loses the sync of my mouse, a synaptic. Any issue ?
>>    
>>
>
>Can you repeat it? Losing sync of mouse is probably a bug in inputs,
>and should be fixed.
>								Pavel
>  
>


[-- Attachment #2: Type: text/html, Size: 1991 bytes --]

Re: SCI interrupt handler : spend too much time

[Pavel Machek]

Hi!

> It sound like serious one, if SCI interrupt handler was blamed for
> this issue.     --Luming

I investigated a bit, and interrupt latency alone is not enough to
break mouse. [Unfortunately, I do not have easy/sure way to reproduce
it.]

Is it possible that SCI interrupt handler accesses keyboard controller
(and breaks mouse?)
								Pavel



-- 
Horseback riding is like software...
...vgf orggre jura vgf serr.


-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive?  Does it
help you create better code?  SHARE THE LOVE, and help us help
YOU!  Click Here: http://sourceforge.net/donate/

Re: SCI interrupt handler : spend too much time

[roychris]

[-- Attachment #1: Type: text/plain, Size: 637 bytes --]

Just a precision,
synaptic driver is not like the other mouses. psmouse call a specific 
fonction for this touchpad that is in an other module.
The problem happens since 2.6, since psmouse and synaptic had been 
seperated....

Pavel Machek wrote:

>Hi!
>
>  
>
>>It sound like serious one, if SCI interrupt handler was blamed for
>>this issue.     --Luming
>>    
>>
>
>I investigated a bit, and interrupt latency alone is not enough to
>break mouse. [Unfortunately, I do not have easy/sure way to reproduce
>it.]
>
>Is it possible that SCI interrupt handler accesses keyboard controller
>(and breaks mouse?)
>								Pavel
>
>
>
>  
>


[-- Attachment #2: Type: text/html, Size: 1069 bytes --]

RE: SCI interrupt handler : spend too much time

[Yu, Luming]

[-- Attachment #1: Type: text/plain, Size: 1046 bytes --]

Does it mean the overhead of SCI interrupt was not the root cause of this issue. --Luming
 

-----Original Message-----
From: roychris [mailto:roychris-GANU6spQydw@public.gmane.org]
Sent: 2003?12?3? 20:49
To: Pavel Machek
Cc: Yu, Luming; acpi-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
Subject: Re: [ACPI] SCI interrupt handler : spend too much time


Just a precision,
synaptic driver is not like the other mouses. psmouse call a specific fonction for this touchpad that is in an other module.
The problem happens since 2.6, since psmouse and synaptic had been seperated....

Pavel Machek wrote:


Hi!



  

It sound like serious one, if SCI interrupt handler was blamed for

this issue.     --Luming

    



I investigated a bit, and interrupt latency alone is not enough to

break mouse. [Unfortunately, I do not have easy/sure way to reproduce

it.]



Is it possible that SCI interrupt handler accesses keyboard controller

(and breaks mouse?)

								Pavel







  



[-- Attachment #2: Type: text/html, Size: 1942 bytes --]

Re: SCI interrupt handler : spend too much time

[roychris]

[-- Attachment #1: Type: text/plain, Size: 1208 bytes --]

Personally, I think so.

Yu, Luming a écrit :

> Does it mean the overhead of SCI interrupt was not the root cause of 
> this issue. --Luming
>  
>
>     -----Original Message-----
>     *From:* roychris [mailto:roychris-GANU6spQydw@public.gmane.org]
>     *Sent:* 2003?12?3? 20:49
>     *To:* Pavel Machek
>     *Cc:* Yu, Luming; acpi-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
>     *Subject:* Re: [ACPI] SCI interrupt handler : spend too much time
>
>     Just a precision,
>     synaptic driver is not like the other mouses. psmouse call a
>     specific fonction for this touchpad that is in an other module.
>     The problem happens since 2.6, since psmouse and synaptic had been
>     seperated....
>
>     Pavel Machek wrote:
>
>>Hi!
>>
>>  
>>
>>>It sound like serious one, if SCI interrupt handler was blamed for
>>>this issue.     --Luming
>>>    
>>>
>>
>>I investigated a bit, and interrupt latency alone is not enough to
>>break mouse. [Unfortunately, I do not have easy/sure way to reproduce
>>it.]
>>
>>Is it possible that SCI interrupt handler accesses keyboard controller
>>(and breaks mouse?)
>>								Pavel
>>
>>
>>
>>  
>>
>


[-- Attachment #2: Type: text/html, Size: 2437 bytes --]

Re: SCI interrupt handler : spend too much time

[Pavel Machek]

Hi!

>    Does it mean the overhead of SCI interrupt was not the root cause of
>    this issue. --Luming

It certainly looks so. Is it possible that SCI handler touches
keyboard controller?

-- 
Horseback riding is like software...
...vgf orggre jura vgf serr.


-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive?  Does it
help you create better code?  SHARE THE LOVE, and help us help
YOU!  Click Here: http://sourceforge.net/donate/

RE: SCI interrupt handler : spend too much time

[Yu, Luming]

To SCI handler itself,  I don't think keyboard controller will be touched. But SCI handler could evaluate some AML method which could touch keyboard controller. --Luming

-----Original Message-----
From: Pavel Machek [mailto:pavel-+ZI9xUNit7I@public.gmane.org]
Sent: 2003?12?4? 17:53
To: Yu, Luming
Cc: roychris; Pavel Machek; acpi-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
Subject: Re: [ACPI] SCI interrupt handler : spend too much time


Hi!

>    Does it mean the overhead of SCI interrupt was not the root cause of
>    this issue. --Luming

It certainly looks so. Is it possible that SCI handler touches
keyboard controller?

-- 
Horseback riding is like software...
...vgf orggre jura vgf serr.


-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive?  Does it
help you create better code?  SHARE THE LOVE, and help us help
YOU!  Click Here: http://sourceforge.net/donate/

Re: SCI interrupt handler : spend too much time

[Pavel Machek]

Hi!

> To SCI handler itself,  I don't think keyboard controller will be touched. But SCI handler could evaluate some AML method which could touch keyboard controller. --Luming
> 

Would that be a bug in AML method? Or how is keyboard-controller
access locked between OS and ACPI?
									Pavel

> >    Does it mean the overhead of SCI interrupt was not the root cause of
> >    this issue. --Luming
> 
> It certainly looks so. Is it possible that SCI handler touches
> keyboard controller?
> 

-- 
When do you have heart between your knees?


-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive?  Does it
help you create better code?  SHARE THE LOVE, and help us help
YOU!  Click Here: http://sourceforge.net/donate/

Re: SCI interrupt handler : spend too much time

[roychris]

[-- Attachment #1: Type: text/plain, Size: 667 bytes --]

Hi,
I have no rule about mouse or keyboard in my aml code.
Pavel Machek a écrit :

>Hi!
>
>  
>
>>To SCI handler itself,  I don't think keyboard controller will be touched. But SCI handler could evaluate some AML method which could touch keyboard controller. --Luming
>>
>>    
>>
>
>Would that be a bug in AML method? Or how is keyboard-controller
>access locked between OS and ACPI?
>									Pavel
>
>  
>
>>>   Does it mean the overhead of SCI interrupt was not the root cause of
>>>   this issue. --Luming
>>>      
>>>
>>It certainly looks so. Is it possible that SCI handler touches
>>keyboard controller?
>>
>>    
>>
>
>  
>


[-- Attachment #2: Type: text/html, Size: 1237 bytes --]

RE: SCI interrupt handler : spend too much time

[Li, Shaohua]

Hi,
As I knew, some systems use keyboard controller to implement EC (I'm not sure, but some materials said). Maybe the error it's caused by EC, so I recommend you to try the patch in http://bugme.osdl.org/show_bug.cgi?id=1433. 

Thanks,
Shaohua


> -----Original Message-----
> From: acpi-devel-admin-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org [mailto:acpi-devel-
> admin-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org] On Behalf Of Yu, Luming
> Sent: 2003年12月8日 11:21
> To: Pavel Machek
> Cc: roychris; acpi-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
> Subject: RE: [ACPI] SCI interrupt handler : spend too much time
> 
> To SCI handler itself,  I don't think keyboard controller will be touched.
> But SCI handler could evaluate some AML method which could touch keyboard
> controller. --Luming
> 
> -----Original Message-----
> From: Pavel Machek [mailto:pavel-+ZI9xUNit7I@public.gmane.org]
> Sent: 2003?12?4? 17:53
> To: Yu, Luming
> Cc: roychris; Pavel Machek; acpi-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
> Subject: Re: [ACPI] SCI interrupt handler : spend too much time
> 
> 
> Hi!
> 
> >    Does it mean the overhead of SCI interrupt was not the root cause of
> >    this issue. --Luming
> 
> It certainly looks so. Is it possible that SCI handler touches
> keyboard controller?
> 
> --
> Horseback riding is like software...
> ...vgf orggre jura vgf serr.
> 
> 
> -------------------------------------------------------
> This SF.net email is sponsored by: SF.net Giveback Program.
> Does SourceForge.net help you be more productive?  Does it
> help you create better code?  SHARE THE LOVE, and help us help
> YOU!  Click Here: http://sourceforge.net/donate/
> _______________________________________________
> Acpi-devel mailing list
> Acpi-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
> https://lists.sourceforge.net/lists/listinfo/acpi-devel


-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive?  Does it
help you create better code?  SHARE THE LOVE, and help us help
YOU!  Click Here: http://sourceforge.net/donate/

Re: SCI interrupt handler : spend too much time

[roychris]

[-- Attachment #1: Type: text/plain, Size: 2592 bytes --]

Hi,
Sorry, it doesn't resolve my problem.

Li, Shaohua a écrit:

>Hi,
>As I knew, some systems use keyboard controller to implement EC (I'm not sure, but some materials said). Maybe the error it's caused by EC, so I recommend you to try the patch in http://bugme.osdl.org/show_bug.cgi?id=1433. 
>
>Thanks,
>Shaohua
>
>
>  
>
>>-----Original Message-----
>>From: acpi-devel-admin-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org [mailto:acpi-devel-
>>admin-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org] On Behalf Of Yu, Luming
>>Sent: 2003年12月8日 11:21
>>To: Pavel Machek
>>Cc: roychris; acpi-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
>>Subject: RE: [ACPI] SCI interrupt handler : spend too much time
>>
>>To SCI handler itself,  I don't think keyboard controller will be touched.
>>But SCI handler could evaluate some AML method which could touch keyboard
>>controller. --Luming
>>
>>-----Original Message-----
>>From: Pavel Machek [mailto:pavel-+ZI9xUNit7I@public.gmane.org]
>>Sent: 2003?12?4? 17:53
>>To: Yu, Luming
>>Cc: roychris; Pavel Machek; acpi-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
>>Subject: Re: [ACPI] SCI interrupt handler : spend too much time
>>
>>
>>Hi!
>>
>>    
>>
>>>   Does it mean the overhead of SCI interrupt was not the root cause of
>>>   this issue. --Luming
>>>      
>>>
>>It certainly looks so. Is it possible that SCI handler touches
>>keyboard controller?
>>
>>--
>>Horseback riding is like software...
>>...vgf orggre jura vgf serr.
>>
>>
>>-------------------------------------------------------
>>This SF.net email is sponsored by: SF.net Giveback Program.
>>Does SourceForge.net help you be more productive?  Does it
>>help you create better code?  SHARE THE LOVE, and help us help
>>YOU!  Click Here: http://sourceforge.net/donate/
>>_______________________________________________
>>Acpi-devel mailing list
>>Acpi-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
>>https://lists.sourceforge.net/lists/listinfo/acpi-devel
>>    
>>
>
>
>-------------------------------------------------------
>This SF.net email is sponsored by: SF.net Giveback Program.
>Does SourceForge.net help you be more productive?  Does it
>help you create better code?  SHARE THE LOVE, and help us help
>YOU!  Click Here: http://sourceforge.net/donate/
>_______________________________________________
>Acpi-devel mailing list
>Acpi-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
>https://lists.sourceforge.net/lists/listinfo/acpi-devel
>
>
>  
>


[-- Attachment #2: Type: text/html, Size: 4256 bytes --]