RE: Processor C states and USB problem

RE: Processor C states and USB problem

[Yu, Luming]

>usb host use poll mode to get data from usb peripherals, and result is
>continuous busmaster activity.

Hmm,  it looks like we should stop such kind of poll to support power
save.






-------------------------------------------------------
This SF.net email is sponsored by: IBM Linux Tutorials.
Become an expert in LINUX or just sharpen your skills.  Sign up for IBM's
Free Linux Tutorials.  Learn everything from the bash shell to sys admin.
Click now! http://ads.osdn.com/?ad_id\x1278&alloc_id371&op=click

RE: Processor C states and USB problem

[Nate Lawson]

On Tue, 6 Jan 2004, Yu, Luming wrote:
> >usb host use poll mode to get data from usb peripherals, and result is
> >continuous busmaster activity.
>
> Hmm,  it looks like we should stop such kind of poll to support power
> save.

Good luck.  That would require processor.c to become aware of the USB
driver loaded.  In FreeBSD, we have a generic kernel interface module (.m)
that you can add that allows other kernel features to call a function
without knowing what will service the call in the backend.  I guess you
could do this via a function pointer, but you'd still have added binary
compatibility issues between ACPI and USB.

-Nate


-------------------------------------------------------
This SF.net email is sponsored by: IBM Linux Tutorials.
Become an expert in LINUX or just sharpen your skills.  Sign up for IBM's
Free Linux Tutorials.  Learn everything from the bash shell to sys admin.
Click now! http://ads.osdn.com/?ad_id=1278&alloc_id=3371&op=click

RE: Processor C states and USB problem

[Nate Lawson]

On Sun, 18 Jan 2004, Yu, Luming wrote:
> > Yes, but compared to the usage count of C2 that is extremely low, and
> > the CPU is idling all the time (CPU usage is below 1%). I don't know
> > much about C states but I guess it should go into C3 much more often,
> > and the bus mastering activity should show something else
> > than ffffffff
> > when booting with a bare bones kernel and with no extra hw initialized
> > (USB, FireWire, etc.)
>
> I agree that it's unreasonable that C3 usage count is never above 5.
> This is the result of the C state power policy.
>
> Threshold is:
> pr->power.states[ACPI_STATE_C2].promotion.threshold.bm = 0x0F;
> pr->power.states[ACPI_STATE_C3].demotion.threshold.bm = 0x0F;
>
> The condition of promotion or demotion is:
> if (pr->power.bm_activity & cx->demotion.threshold.bm) {
> if (!(pr->power.bm_activity & cx->promotion.threshold.bm)) {
>
> Using the following statement to adjust bm_activity of the last Nth
> schedule quanta.
>
> pr->power.bm_activity <<= 1;
> if(bm_status)
> 	pr->power.bm_activity++;
>
> So it's is not strange that C3 usage count is 4  !
>
> Maybe we need a more sophisticated policy .

In FreeBSD, I clear BM_STS before exiting idle.  If I come back and it's
set, then avoid C3 and demote to C2.  If not and last sleep was > C3
latency, then use C3.  It's unnecessary to count the number of BM_STS
events.

-Nate


-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004
Premiere Conference on Open Tools Development and Integration
See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn

RE: Processor C states and USB problem

[Diego SANTA CRUZ]

On Sun, 2004-01-18 at 04:17, Yu, Luming wrote:
> > 
> > Yes, but compared to the usage count of C2 that is extremely low, and
> > the CPU is idling all the time (CPU usage is below 1%). I don't know
> > much about C states but I guess it should go into C3 much more often,
> > and the bus mastering activity should show something else 
> > than ffffffff
> > when booting with a bare bones kernel and with no extra hw initialized
> > (USB, FireWire, etc.)
> 
> I agree that it's unreasonable that C3 usage count is never above 5. 
> This is the result of the C state power policy. 

Thanks for the explanations, it is more or less what I had understood
from reading processor.c last week. But my bottom line question was more
of the sort "why is there constant bus mastering activity if almost no
hardware is initialized", as I understand it is the bus mastering
activity that's preventing the use of C3.

Is there a not-so-complicated way to find out which subsystem is the
culprit for the bus mastering activity? As reported elsewhere in this
thread, it appears to be that this issue appears under 2.4 and not 2.6
(provided no USB mouse is plugged). I have not tested 2.6 on my laptop
yet, only 2.4.

Best,

Diego

> 
> Threshold is:
> pr->power.states[ACPI_STATE_C2].promotion.threshold.bm = 0x0F;
> pr->power.states[ACPI_STATE_C3].demotion.threshold.bm = 0x0F;
> 
> The condition of promotion or demotion is:
> if (pr->power.bm_activity & cx->demotion.threshold.bm) {
> if (!(pr->power.bm_activity & cx->promotion.threshold.bm)) {
> 
> Using the following statement to adjust bm_activity of the last Nth
> schedule quanta.
> 
> pr->power.bm_activity <<= 1;
> if(bm_status)
> 	pr->power.bm_activity++;
> 
> So it's is not strange that C3 usage count is 4  !
> 
> Maybe we need a more sophisticated policy .
> 
> 
> 
> 
> 
> 
> 
> -------------------------------------------------------
> The SF.Net email is sponsored by EclipseCon 2004
> Premiere Conference on Open Tools Development and Integration
> See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
> http://www.eclipsecon.org/osdn
> _______________________________________________
> Acpi-devel mailing list
> Acpi-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
> https://lists.sourceforge.net/lists/listinfo/acpi-devel
-- 
Diego SANTA CRUZ <dsanta-W/9V78bTXrinCNX9eg9BBA@public.gmane.org>
Ecole Polytechnique Fédérale de Lausanne - EPFL



-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004
Premiere Conference on Open Tools Development and Integration
See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn

RE: Processor C states and USB problem

[Ow Mun Heng]

> -----Original Message-----
> From: Yu, Luming [mailto:luming.yu-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org]
> > Under 2.4 bus master always shows "ffffffff" on my 
> notebook. Sometimes
> > shutting down USB helps, sometimes it doesn't.
> > Running under 2.6 I only get full bus master activity if an USB
> > mouse is plugged in.
> 
> Could you post output from " cat /proc/acpi/processor/CPU0/power"
> I want to check the usage of C3.

RH9 + 2.4.24 + cpufreq patched + D600
No USB attached. USB kernel module is loaded.. I can rmmod all but usbcore
(still ffffff)

 # cat /proc/acpi/processor/CPU0/power
 active state:            C2
 default state:           C1
 bus master activity:     ffffffff
 states:
     C1:                  promotion[C2] demotion[--] 
 latency[000] usage[00212900]
    *C2:                  promotion[C3] demotion[C1] 
 latency[050] usage[02242678]
     C3:                  promotion[--] demotion[C2] 
 latency[050] usage[00000000]




-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004
Premiere Conference on Open Tools Development and Integration
See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn

Re: Processor C states and USB problem

[Michael Guntsche]

On Sun, 18 Jan 2004 13:57:59 +0800
"Yu, Luming" <luming.yu-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org> wrote:

> 
> > Under 2.4 bus master always shows "ffffffff" on my notebook. Sometimes
> > shutting down USB helps, sometimes it doesn't.
> > Running under 2.6 I only get full bus master activity if an USB
> > mouse is plugged in.
> 
> Could you post output from " cat /proc/acpi/processor/CPU0/power"
> I want to check the usage of C3.

active state:            C2
default state:           C1
bus master activity:     9292d292
states:
    C1:                  promotion[C2] demotion[--] latency[000] usage[00000010]
   *C2:                  promotion[C3] demotion[C1] latency[001] usage[00086443]
    C3:                  promotion[--] demotion[C2] latency[085] usage[00012776]


This is 2.6.1 with no USB mouse plugged in.


Michael


-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004
Premiere Conference on Open Tools Development and Integration
See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn

RE: Processor C states and USB problem

[Yu, Luming]

> Under 2.4 bus master always shows "ffffffff" on my notebook. Sometimes
> shutting down USB helps, sometimes it doesn't.
> Running under 2.6 I only get full bus master activity if an USB
> mouse is plugged in.

Could you post output from " cat /proc/acpi/processor/CPU0/power"
I want to check the usage of C3.


-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004
Premiere Conference on Open Tools Development and Integration
See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn

RE: Processor C states and USB problem

[Yu, Luming]

> 
> Yes, but compared to the usage count of C2 that is extremely low, and
> the CPU is idling all the time (CPU usage is below 1%). I don't know
> much about C states but I guess it should go into C3 much more often,
> and the bus mastering activity should show something else 
> than ffffffff
> when booting with a bare bones kernel and with no extra hw initialized
> (USB, FireWire, etc.)

I agree that it's unreasonable that C3 usage count is never above 5. 
This is the result of the C state power policy. 

Threshold is:
pr->power.states[ACPI_STATE_C2].promotion.threshold.bm = 0x0F;
pr->power.states[ACPI_STATE_C3].demotion.threshold.bm = 0x0F;

The condition of promotion or demotion is:
if (pr->power.bm_activity & cx->demotion.threshold.bm) {
if (!(pr->power.bm_activity & cx->promotion.threshold.bm)) {

Using the following statement to adjust bm_activity of the last Nth
schedule quanta.

pr->power.bm_activity <<= 1;
if(bm_status)
	pr->power.bm_activity++;

So it's is not strange that C3 usage count is 4  !

Maybe we need a more sophisticated policy .







-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004
Premiere Conference on Open Tools Development and Integration
See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn

Re: Processor C states and USB problem

[Diego SANTA CRUZ]

On Wed, 2004-01-14 at 16:34, Michael Guntsche wrote:

> 
> Have you tried running a 2.6 kernel?
> Under 2.4 bus master always shows "ffffffff" on my notebook. Sometimes
> shutting down USB helps, sometimes it doesn't.
> Running under 2.6 I only get full bus master activity if an USB
> mouse is plugged in.
> 

Thanks for the info. I guess I should get myself to try it out sometime
soon then. The only problem is that the winmodem in this laptop uses a
binary driver not available for 2.6 (AFAIK).

Best,

Diego

-- 
-------------------------------------------------------
Diego Santa Cruz - PhD
Publications available at http://ltswww.epfl.ch/~dsanta
Signal Processing Institute (LTS1 / ITS / STI)
Swiss Federal Institute of Technology (EPFL)
EPFL - STI - ITS, CH-1015 Lausanne, Switzerland
E-mail:     Diego.SantaCruz-p8DiymsW2f8@public.gmane.org
Phone:      +41 - 21 - 693 26 57
Fax:        +41 - 21 - 693 76 00
-------------------------------------------------------



-------------------------------------------------------
This SF.net email is sponsored by: Perforce Software.
Perforce is the Fast Software Configuration Management System offering
advanced branching capabilities and atomic changes on 50+ platforms.
Free Eval! http://www.perforce.com/perforce/loadprog.html

Re: Processor C states and USB problem

[Michael Guntsche]

On Wed, 14 Jan 2004 08:55:35 +0100
Diego SANTA CRUZ <dsanta-W/9V78bTXrinCNX9eg9BBA@public.gmane.org> wrote:

> On Wed, 2004-01-14 at 03:10, Yu, Luming wrote:
> > > > > I picked up this thread and I'm interested in knowing a 
> > > bit more about
> > > > > bus mastering and C3, as my laptop never goes into C3. I 
> > > own a Dell
> > > > > Inspiron 8500 (details at
> > > > > http://ltswww.epfl.ch/~dsanta/resources/dell-i8500-linux).
> > > > > 
> > 
> > How did you know your processor never transition to C3 mode?
> > 
> 
> The usage count is never above 5, even after a long period (an hour or
> more of idle CPU). But yes, strictly speaking it's almost never ;-)
> 
> > > > > This is what I get for the power states:
> > > > > # cat /proc/acpi/processor/CPU0/power
> > > > > active state:            C2
> > > > > default state:           C1
> > > > > bus master activity:     ffffffff
> > > > > states:
> > > > >     C1:                  promotion[C2] demotion[--] 
> > > > > latency[000] usage[00039800]
> > > > >    *C2:                  promotion[C3] demotion[C1] 
> > > > > latency[050] usage[00573330]
> > > > >     C3:                  promotion[--] demotion[C2] 
> > > > > latency[050] usage[00000004]
> > > > > 
> > 
> > Usage of C3 is 4. 
> 
> Yes, but compared to the usage count of C2 that is extremely low, and
> the CPU is idling all the time (CPU usage is below 1%). I don't know
> much about C states but I guess it should go into C3 much more often,
> and the bus mastering activity should show something else than
> ffffffff when booting with a bare bones kernel and with no extra hw
> initialized(USB, FireWire, etc.)
> 

Have you tried running a 2.6 kernel?
Under 2.4 bus master always shows "ffffffff" on my notebook. Sometimes
shutting down USB helps, sometimes it doesn't.
Running under 2.6 I only get full bus master activity if an USB
mouse is plugged in.


/michael


-------------------------------------------------------
This SF.net email is sponsored by: Perforce Software.
Perforce is the Fast Software Configuration Management System offering
advanced branching capabilities and atomic changes on 50+ platforms.
Free Eval! http://www.perforce.com/perforce/loadprog.html

RE: Processor C states and USB problem

[Diego SANTA CRUZ]

On Wed, 2004-01-14 at 03:10, Yu, Luming wrote:
> > > > I picked up this thread and I'm interested in knowing a 
> > bit more about
> > > > bus mastering and C3, as my laptop never goes into C3. I 
> > own a Dell
> > > > Inspiron 8500 (details at
> > > > http://ltswww.epfl.ch/~dsanta/resources/dell-i8500-linux).
> > > > 
> 
> How did you know your processor never transition to C3 mode?
> 

The usage count is never above 5, even after a long period (an hour or
more of idle CPU). But yes, strictly speaking it's almost never ;-)

> > > > This is what I get for the power states:
> > > > # cat /proc/acpi/processor/CPU0/power
> > > > active state:            C2
> > > > default state:           C1
> > > > bus master activity:     ffffffff
> > > > states:
> > > >     C1:                  promotion[C2] demotion[--] 
> > > > latency[000] usage[00039800]
> > > >    *C2:                  promotion[C3] demotion[C1] 
> > > > latency[050] usage[00573330]
> > > >     C3:                  promotion[--] demotion[C2] 
> > > > latency[050] usage[00000004]
> > > > 
> 
> Usage of C3 is 4. 

Yes, but compared to the usage count of C2 that is extremely low, and
the CPU is idling all the time (CPU usage is below 1%). I don't know
much about C states but I guess it should go into C3 much more often,
and the bus mastering activity should show something else than ffffffff
when booting with a bare bones kernel and with no extra hw initialized
(USB, FireWire, etc.)

Best,

Diego

-- 
Diego SANTA CRUZ <dsanta-W/9V78bTXrinCNX9eg9BBA@public.gmane.org>
Ecole Polytechnique Fédérale de Lausanne - EPFL



-------------------------------------------------------
This SF.net email is sponsored by: Perforce Software.
Perforce is the Fast Software Configuration Management System offering
advanced branching capabilities and atomic changes on 50+ platforms.
Free Eval! http://www.perforce.com/perforce/loadprog.html

RE: Processor C states and USB problem

[Yu, Luming]

> > > I picked up this thread and I'm interested in knowing a 
> bit more about
> > > bus mastering and C3, as my laptop never goes into C3. I 
> own a Dell
> > > Inspiron 8500 (details at
> > > http://ltswww.epfl.ch/~dsanta/resources/dell-i8500-linux).
> > > 

How did you know your processor never transition to C3 mode?

> > > This is what I get for the power states:
> > > # cat /proc/acpi/processor/CPU0/power
> > > active state:            C2
> > > default state:           C1
> > > bus master activity:     ffffffff
> > > states:
> > >     C1:                  promotion[C2] demotion[--] 
> > > latency[000] usage[00039800]
> > >    *C2:                  promotion[C3] demotion[C1] 
> > > latency[050] usage[00573330]
> > >     C3:                  promotion[--] demotion[C2] 
> > > latency[050] usage[00000004]
> > > 

Usage of C3 is 4. 



-------------------------------------------------------
This SF.net email is sponsored by: Perforce Software.
Perforce is the Fast Software Configuration Management System offering
advanced branching capabilities and atomic changes on 50+ platforms.
Free Eval! http://www.perforce.com/perforce/loadprog.html

RE: Processor C states and USB problem

[Diego SANTA CRUZ]

On Mon, 2004-01-12 at 09:31, Yu, Luming wrote:
> > I picked up this thread and I'm interested in knowing a bit more about
> > bus mastering and C3, as my laptop never goes into C3. I own a Dell
> > Inspiron 8500 (details at
> > http://ltswww.epfl.ch/~dsanta/resources/dell-i8500-linux).
> > 
> > This is what I get for the power states:
> > # cat /proc/acpi/processor/CPU0/power
> > active state:            C2
> > default state:           C1
> > bus master activity:     ffffffff
> > states:
> >     C1:                  promotion[C2] demotion[--] 
> > latency[000] usage[00039800]
> >    *C2:                  promotion[C3] demotion[C1] 
> > latency[050] usage[00573330]
> >     C3:                  promotion[--] demotion[C2] 
> > latency[050] usage[00000004]
> > 
> > I tried booting with init=/bin/sh and manually loading the processor
> > module, without any external hardware attached, so that no special
> > hardware is initialized, but still there is non-stopping bus mastering
> > activity. What could be causing it? Any ideas how could I find out?
> 
> I just filed a tracker http://bugme.osdl.org/show_bug.cgi?id=1839.
> 

I tested with a bare bones kernel config and the problem persists. I
entered all the data that I thought useful in the bugzilla entry above.
Let me know if anything else is needed to help debug the problem.

Best,

Diego
-- 
-------------------------------------------------------
Diego Santa Cruz - PhD
Publications available at http://ltswww.epfl.ch/~dsanta
Signal Processing Institute (LTS1 / ITS / STI)
Swiss Federal Institute of Technology (EPFL)
EPFL - STI - ITS, CH-1015 Lausanne, Switzerland
E-mail:     Diego.SantaCruz-p8DiymsW2f8@public.gmane.org
Phone:      +41 - 21 - 693 26 57
Fax:        +41 - 21 - 693 76 00
-------------------------------------------------------



-------------------------------------------------------
This SF.net email is sponsored by: Perforce Software.
Perforce is the Fast Software Configuration Management System offering
advanced branching capabilities and atomic changes on 50+ platforms.
Free Eval! http://www.perforce.com/perforce/loadprog.html

RE: Processor C states and USB problem

[Yu, Luming]

> I picked up this thread and I'm interested in knowing a bit more about
> bus mastering and C3, as my laptop never goes into C3. I own a Dell
> Inspiron 8500 (details at
> http://ltswww.epfl.ch/~dsanta/resources/dell-i8500-linux).
> 
> This is what I get for the power states:
> # cat /proc/acpi/processor/CPU0/power
> active state:            C2
> default state:           C1
> bus master activity:     ffffffff
> states:
>     C1:                  promotion[C2] demotion[--] 
> latency[000] usage[00039800]
>    *C2:                  promotion[C3] demotion[C1] 
> latency[050] usage[00573330]
>     C3:                  promotion[--] demotion[C2] 
> latency[050] usage[00000004]
> 
> I tried booting with init=/bin/sh and manually loading the processor
> module, without any external hardware attached, so that no special
> hardware is initialized, but still there is non-stopping bus mastering
> activity. What could be causing it? Any ideas how could I find out?

I just filed a tracker http://bugme.osdl.org/show_bug.cgi?id=1839.


-------------------------------------------------------
This SF.net email is sponsored by: Perforce Software.
Perforce is the Fast Software Configuration Management System offering
advanced branching capabilities and atomic changes on 50+ platforms.
Free Eval! http://www.perforce.com/perforce/loadprog.html

Re: Processor C states and USB problem

[Michael Guntsche]

On Fri, 9 Jan 2004 09:35:53 -0800
"Randy.Dunlap" <rddunlap-3NddpPZAyC0@public.gmane.org> wrote:

> On Fri, 9 Jan 2004 16:54:21 +0100 Michael Guntsche <mike-Z92qn3yYq0hWk0Htik3J/w@public.gmane.org>
> wrote:
> 
> | On Fri, 09 Jan 2004 11:15:22 +0100
> | Diego SANTA CRUZ <Diego.SantaCruz-p8DiymsW2f8@public.gmane.org> wrote:
> | 
> | > On Fri, 2004-01-09 at 11:05, Petr Olivka wrote:
> | > > in 2.4 kernel I have to unload all usb modules. Then power drop
> | > > to C3 state. If I remember, only usbcore can be loaded (on my
> | > > D600).
> | > > 
> | > 
> | > But even booting with init=/bin/sh and without ever loading any
> | > USB module I get permanent bus mastering activity. Any ideas?
> | 
> | I see the same on my notebook with a 2.4 kernel. Sometimes it works,
> | but most of the time there is Bus master acitivty.
> | With 2.6 I only see that if I have a USB device plugged in (mouse)
> | that gets polled.
> 
> It seems like one of the early usb uhci host controller drivers (HCDs)
> generated lots of bus master activity.  Are you sure that uhci or
> usb-uhci is not loaded (in the 2.4 problem case), even though no
> USB device drivers are loaded?
> 
Hmm, I did some tests and found out the following.

With USB modules loaded

	bus master activity:     ffffffff

With USB modules removed, the number goes down, but the processor still
stays in C2 most of the time.


In 2.6 I only get "ffffffff" if the mouse is plugged otherwise the
notebook spends more time in C3.
Since I am using 2.6 nowadays (PREEMPT being one of the main reasons),
this is not very important for me.

On a side note.
Do you have an idea why I no longer have performance management and the limit interface with 2.6?
I don't use those two myself, I just noticed it while testing ACPI in 2.6. 

/proc/acpi/processor/CPU0/info:
processor id:            0
acpi id:                 0
bus mastering control:   yes
power management:        yes
throttling control:      no
performance management:  no <-- yes on 2.4
limit interface:         no <-- yes on 2.4

This is an Acer 800 notebook with a Pentium M processor.

/Michael


 


-------------------------------------------------------
This SF.net email is sponsored by: Perforce Software.
Perforce is the Fast Software Configuration Management System offering
advanced branching capabilities and atomic changes on 50+ platforms.
Free Eval! http://www.perforce.com/perforce/loadprog.html

Re: Processor C states and USB problem

[Randy.Dunlap]

On Fri, 9 Jan 2004 16:54:21 +0100 Michael Guntsche <mike-Z92qn3yYq0hWk0Htik3J/w@public.gmane.org> wrote:

| On Fri, 09 Jan 2004 11:15:22 +0100
| Diego SANTA CRUZ <Diego.SantaCruz-p8DiymsW2f8@public.gmane.org> wrote:
| 
| > On Fri, 2004-01-09 at 11:05, Petr Olivka wrote:
| > > in 2.4 kernel I have to unload all usb modules. Then power drop to C3
| > > state. If I remember, only usbcore can be loaded (on my D600).
| > > 
| > 
| > But even booting with init=/bin/sh and without ever loading any USB
| > module I get permanent bus mastering activity. Any ideas?
| 
| I see the same on my notebook with a 2.4 kernel. Sometimes it works, but most of the time
| there is Bus master acitivty.
| With 2.6 I only see that if I have a USB device plugged in (mouse) that gets polled.

It seems like one of the early usb uhci host controller drivers (HCDs)
generated lots of bus master activity.  Are you sure that uhci or
usb-uhci is not loaded (in the 2.4 problem case), even though no
USB device drivers are loaded?

--
~Randy
MOTD:  Always include version info.


-------------------------------------------------------
This SF.net email is sponsored by: Perforce Software.
Perforce is the Fast Software Configuration Management System offering
advanced branching capabilities and atomic changes on 50+ platforms.
Free Eval! http://www.perforce.com/perforce/loadprog.html

Re: Processor C states and USB problem

[Michael Guntsche]

On Fri, 09 Jan 2004 11:15:22 +0100
Diego SANTA CRUZ <Diego.SantaCruz-p8DiymsW2f8@public.gmane.org> wrote:

> On Fri, 2004-01-09 at 11:05, Petr Olivka wrote:
> > in 2.4 kernel I have to unload all usb modules. Then power drop to C3
> > state. If I remember, only usbcore can be loaded (on my D600).
> > 
> 
> But even booting with init=/bin/sh and without ever loading any USB
> module I get permanent bus mastering activity. Any ideas?

I see the same on my notebook with a 2.4 kernel. Sometimes it works, but most of the time
there is Bus master acitivty.
With 2.6 I only see that if I have a USB device plugged in (mouse) that gets polled.


Michael


-------------------------------------------------------
This SF.net email is sponsored by: Perforce Software.
Perforce is the Fast Software Configuration Management System offering
advanced branching capabilities and atomic changes on 50+ platforms.
Free Eval! http://www.perforce.com/perforce/loadprog.html

RE: Processor C states and USB problem

[Diego SANTA CRUZ]

On Fri, 2004-01-09 at 11:05, Petr Olivka wrote:
> in 2.4 kernel I have to unload all usb modules. Then power drop to C3
> state. If I remember, only usbcore can be loaded (on my D600).
> 

But even booting with init=/bin/sh and without ever loading any USB
module I get permanent bus mastering activity. Any ideas?

Diego

> poli
> 
> 
> On Fri, 9 Jan 2004, Diego SANTA CRUZ wrote:
> 
> > Hi,
> >
> > I picked up this thread and I'm interested in knowing a bit more about
> > bus mastering and C3, as my laptop never goes into C3. I own a Dell
> > Inspiron 8500 (details at
> > http://ltswww.epfl.ch/~dsanta/resources/dell-i8500-linux).
> >
> > This is what I get for the power states:
> > # cat /proc/acpi/processor/CPU0/power
> > active state:            C2
> > default state:           C1
> > bus master activity:     ffffffff
> > states:
> >     C1:                  promotion[C2] demotion[--] latency[000] usage[00039800]
> >    *C2:                  promotion[C3] demotion[C1] latency[050] usage[00573330]
> >     C3:                  promotion[--] demotion[C2] latency[050] usage[00000004]
> >
> > I tried booting with init=/bin/sh and manually loading the processor
> > module, without any external hardware attached, so that no special
> > hardware is initialized, but still there is non-stopping bus mastering
> > activity. What could be causing it? Any ideas how could I find out?
> >
> > BTW, I'm running the latest Fedora Core 1 kernel (2.4.22-1.2140.nptl)
> >
> > Thanks for any info,
> >
> > Diego
> >
> > On Mon, 2004-01-05 at 10:06, Li, Shaohua wrote:
> > > Hi,
> > > C3 is a kind of idle, it means deep power saving.
> > > Actually, I observed the same symptom as you: USB devices caused many bus master activities, which make processor can't enter C3.
> > >
> > > Thanks,
> > > Shaohua
> >
> > --
> > -------------------------------------------------------
> > Diego Santa Cruz - PhD
> > Publications available at http://ltswww.epfl.ch/~dsanta
> > Signal Processing Institute (LTS1 / ITS / STI)
> > Swiss Federal Institute of Technology (EPFL)
> > EPFL - STI - ITS, CH-1015 Lausanne, Switzerland
> > E-mail:     Diego.SantaCruz-p8DiymsW2f8@public.gmane.org
> > Phone:      +41 - 21 - 693 26 57
> > Fax:        +41 - 21 - 693 76 00
> > -------------------------------------------------------
> >
> >
> >
> > -------------------------------------------------------
> > This SF.net email is sponsored by: Perforce Software.
> > Perforce is the Fast Software Configuration Management System offering
> > advanced branching capabilities and atomic changes on 50+ platforms.
> > Free Eval! http://www.perforce.com/perforce/loadprog.html
> > _______________________________________________
> > Acpi-devel mailing list
> > Acpi-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
> > https://lists.sourceforge.net/lists/listinfo/acpi-devel
> >
-- 
-------------------------------------------------------
Diego Santa Cruz - PhD
Publications available at http://ltswww.epfl.ch/~dsanta
Signal Processing Institute (LTS1 / ITS / STI)
Swiss Federal Institute of Technology (EPFL)
EPFL - STI - ITS, CH-1015 Lausanne, Switzerland
E-mail:     Diego.SantaCruz-p8DiymsW2f8@public.gmane.org
Phone:      +41 - 21 - 693 26 57
Fax:        +41 - 21 - 693 76 00
-------------------------------------------------------



-------------------------------------------------------
This SF.net email is sponsored by: Perforce Software.
Perforce is the Fast Software Configuration Management System offering
advanced branching capabilities and atomic changes on 50+ platforms.
Free Eval! http://www.perforce.com/perforce/loadprog.html

RE: Processor C states and USB problem

[Petr Olivka]

in 2.4 kernel I have to unload all usb modules. Then power drop to C3
state. If I remember, only usbcore can be loaded (on my D600).

poli


On Fri, 9 Jan 2004, Diego SANTA CRUZ wrote:

> Hi,
>
> I picked up this thread and I'm interested in knowing a bit more about
> bus mastering and C3, as my laptop never goes into C3. I own a Dell
> Inspiron 8500 (details at
> http://ltswww.epfl.ch/~dsanta/resources/dell-i8500-linux).
>
> This is what I get for the power states:
> # cat /proc/acpi/processor/CPU0/power
> active state:            C2
> default state:           C1
> bus master activity:     ffffffff
> states:
>     C1:                  promotion[C2] demotion[--] latency[000] usage[00039800]
>    *C2:                  promotion[C3] demotion[C1] latency[050] usage[00573330]
>     C3:                  promotion[--] demotion[C2] latency[050] usage[00000004]
>
> I tried booting with init=/bin/sh and manually loading the processor
> module, without any external hardware attached, so that no special
> hardware is initialized, but still there is non-stopping bus mastering
> activity. What could be causing it? Any ideas how could I find out?
>
> BTW, I'm running the latest Fedora Core 1 kernel (2.4.22-1.2140.nptl)
>
> Thanks for any info,
>
> Diego
>
> On Mon, 2004-01-05 at 10:06, Li, Shaohua wrote:
> > Hi,
> > C3 is a kind of idle, it means deep power saving.
> > Actually, I observed the same symptom as you: USB devices caused many bus master activities, which make processor can't enter C3.
> >
> > Thanks,
> > Shaohua
>
> --
> -------------------------------------------------------
> Diego Santa Cruz - PhD
> Publications available at http://ltswww.epfl.ch/~dsanta
> Signal Processing Institute (LTS1 / ITS / STI)
> Swiss Federal Institute of Technology (EPFL)
> EPFL - STI - ITS, CH-1015 Lausanne, Switzerland
> E-mail:     Diego.SantaCruz-p8DiymsW2f8@public.gmane.org
> Phone:      +41 - 21 - 693 26 57
> Fax:        +41 - 21 - 693 76 00
> -------------------------------------------------------
>
>
>
> -------------------------------------------------------
> This SF.net email is sponsored by: Perforce Software.
> Perforce is the Fast Software Configuration Management System offering
> advanced branching capabilities and atomic changes on 50+ platforms.
> Free Eval! http://www.perforce.com/perforce/loadprog.html
> _______________________________________________
> 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: Perforce Software.
Perforce is the Fast Software Configuration Management System offering
advanced branching capabilities and atomic changes on 50+ platforms.
Free Eval! http://www.perforce.com/perforce/loadprog.html

RE: Processor C states and USB problem

[Diego SANTA CRUZ]

Hi,

I picked up this thread and I'm interested in knowing a bit more about
bus mastering and C3, as my laptop never goes into C3. I own a Dell
Inspiron 8500 (details at
http://ltswww.epfl.ch/~dsanta/resources/dell-i8500-linux).

This is what I get for the power states:
# cat /proc/acpi/processor/CPU0/power
active state:            C2
default state:           C1
bus master activity:     ffffffff
states:
    C1:                  promotion[C2] demotion[--] latency[000] usage[00039800]
   *C2:                  promotion[C3] demotion[C1] latency[050] usage[00573330]
    C3:                  promotion[--] demotion[C2] latency[050] usage[00000004]

I tried booting with init=/bin/sh and manually loading the processor
module, without any external hardware attached, so that no special
hardware is initialized, but still there is non-stopping bus mastering
activity. What could be causing it? Any ideas how could I find out?

BTW, I'm running the latest Fedora Core 1 kernel (2.4.22-1.2140.nptl)

Thanks for any info,

Diego

On Mon, 2004-01-05 at 10:06, Li, Shaohua wrote:
> Hi,
> C3 is a kind of idle, it means deep power saving.
> Actually, I observed the same symptom as you: USB devices caused many bus master activities, which make processor can't enter C3.
>  
> Thanks,
> Shaohua

-- 
-------------------------------------------------------
Diego Santa Cruz - PhD
Publications available at http://ltswww.epfl.ch/~dsanta
Signal Processing Institute (LTS1 / ITS / STI)
Swiss Federal Institute of Technology (EPFL)
EPFL - STI - ITS, CH-1015 Lausanne, Switzerland
E-mail:     Diego.SantaCruz-p8DiymsW2f8@public.gmane.org
Phone:      +41 - 21 - 693 26 57
Fax:        +41 - 21 - 693 76 00
-------------------------------------------------------



-------------------------------------------------------
This SF.net email is sponsored by: Perforce Software.
Perforce is the Fast Software Configuration Management System offering
advanced branching capabilities and atomic changes on 50+ platforms.
Free Eval! http://www.perforce.com/perforce/loadprog.html

RE: Processor C states and USB problem

[Yu, Luming]

>> >usb host use poll mode to get data from usb peripherals, and result
is
>> >continuous busmaster activity.
>>
>> Hmm,  it looks like we should stop such kind of poll to support power
>> save.
>
>Good luck.  That would require processor.c to become aware of the USB
>driver loaded.
>  In FreeBSD, we have a generic kernel interface module (.m)
>that you can add that allows other kernel features to call a function
>without knowing what will service the call in the backend.  I guess you
>could do this via a function pointer, but you'd still have added binary
compatibility issues between ACPI and USB.


Maybe we can do it through a generic driver (PM) .


-------------------------------------------------------
This SF.net email is sponsored by: IBM Linux Tutorials.
Become an expert in LINUX or just sharpen your skills.  Sign up for IBM's
Free Linux Tutorials.  Learn everything from the bash shell to sys admin.
Click now! http://ads.osdn.com/?ad_id\x1278&alloc_id371&op=click

Re: Processor C states and USB problem

[mike-Z92qn3yYq0hWk0Htik3J/w]

Li, Shaohua writes: 

> 
> Hi,
> C3 is a kind of idle, it means deep power saving.
> Actually, I observed the same symptom as you: USB devices caused many bus master activities, which make processor can't enter C3.
>  

So C3 actually IS a good thing. ;)
I somehow got it the wrong way around (C3 being no powersave, C1 being 
powersave).
As for polling the USB mouse causing busmaster acitivty, this makes sense as 
well. 

Thank you two for that prompt answer,
Michael


-------------------------------------------------------
This SF.net email is sponsored by: IBM Linux Tutorials.
Become an expert in LINUX or just sharpen your skills.  Sign up for IBM's
Free Linux Tutorials.  Learn everything from the bash shell to sys admin.
Click now! http://ads.osdn.com/?ad_id=1278&alloc_id=3371&op=click

Re: Processor C states and USB problem

[Petr Olivka]

I did ask about this problem in usb-devel-forum.

You can see answer:
http://sourceforge.net/mailarchive/message.php?msg_id=6249537

usb host use poll mode to get data from usb peripherals, and result is
continuous busmaster activity.

poli


On Mon, 5 Jan 2004 mike-Z92qn3yYq0hWk0Htik3J/w@public.gmane.org wrote:

> Recently I had problems with my network card and ACPI (see earlier posts for
> the problem and solution) so I thought I'd take a closer look at ACPI on
> this notebook (Acer Travelmate 800).
>
> Looking at the proc entries for the processor I saw the following.
>
> active state:            C3
> default state:           C1
> bus master activity:     80070020
> states:
>    C1:                  promotion[C2] demotion[--] latency[000]
> usage[00000010]
>    C2:                  promotion[C3] demotion[C1]
> latency[001]usage[01077887]
>   *C3:                  promotion[--] demotion[C2] latency[085]
> usage[00083507]
>
> This looked pretty normal to me, apart from the fact that the processor was
> spending a lot of time in C3 although the notebook was idle.
>
> Now I plugged in my USB mouse.
>
> cat /proc/acpi/processor/CPU0/power
> active state:            C2
> default state:           C1
> bus master activity:     ffffffff <---- USB is active
> states:
>    C1:                  promotion[C2] demotion[--] latency[000]
> usage[00000010]
>   *C2:                  promotion[C3] demotion[C1]
> latency[001]usage[00871647]
>    C3:                  promotion[--] demotion[C2] latency[085]
> usage[00077386]
>
> The processor never left the C2 state.
> After unplugging the mouse "bus master activity" showed normal values again.
>
> Now for my two questions:
>
> 1.) Is it normal for the notebook to spend so much time in C3, even if it is
> idle?
>
> 2.) Why is an USB device screwing up the bus master activity, resulting in
> the processor never leaving C2?
>
>
> /Michael
>
>
>
>
> -------------------------------------------------------
> This SF.net email is sponsored by: IBM Linux Tutorials.
> Become an expert in LINUX or just sharpen your skills.  Sign up for IBM's
> Free Linux Tutorials.  Learn everything from the bash shell to sys admin.
> Click now! http://ads.osdn.com/?ad_id=1278&alloc_id=3371&op=click
> _______________________________________________
> 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: IBM Linux Tutorials.
Become an expert in LINUX or just sharpen your skills.  Sign up for IBM's
Free Linux Tutorials.  Learn everything from the bash shell to sys admin.
Click now! http://ads.osdn.com/?ad_id=1278&alloc_id=3371&op=click

RE: Processor C states and USB problem

[Li, Shaohua]

Hi,
C3 is a kind of idle, it means deep power saving.
Actually, I observed the same symptom as you: USB devices caused many bus master activities, which make processor can't enter C3.
 
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 mike@it-loops.com
> Sent: 2004年1月5日 16:42
> To: acpi-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
> Subject: [ACPI] Processor C states and USB problem
> 
> Recently I had problems with my network card and ACPI (see earlier posts
> for
> the problem and solution) so I thought I'd take a closer look at ACPI on
> this notebook (Acer Travelmate 800).
> 
> Looking at the proc entries for the processor I saw the following.
> 
> active state:            C3
> default state:           C1
> bus master activity:     80070020
> states:
>    C1:                  promotion[C2] demotion[--] latency[000]
> usage[00000010]
>    C2:                  promotion[C3] demotion[C1]
> latency[001]usage[01077887]
>   *C3:                  promotion[--] demotion[C2] latency[085]
> usage[00083507]
> 
> This looked pretty normal to me, apart from the fact that the processor
> was
> spending a lot of time in C3 although the notebook was idle.
> 
> Now I plugged in my USB mouse.
> 
> cat /proc/acpi/processor/CPU0/power
> active state:            C2
> default state:           C1
> bus master activity:     ffffffff <---- USB is active
> states:
>    C1:                  promotion[C2] demotion[--] latency[000]
> usage[00000010]
>   *C2:                  promotion[C3] demotion[C1]
> latency[001]usage[00871647]
>    C3:                  promotion[--] demotion[C2] latency[085]
> usage[00077386]
> 
> The processor never left the C2 state.
> After unplugging the mouse "bus master activity" showed normal values
> again.
> 
> Now for my two questions:
> 
> 1.) Is it normal for the notebook to spend so much time in C3, even if it
> is
> idle?
> 
> 2.) Why is an USB device screwing up the bus master activity, resulting in
> the processor never leaving C2?
> 
> 
> /Michael
> 
> 
> 
> 
> -------------------------------------------------------
> This SF.net email is sponsored by: IBM Linux Tutorials.
> Become an expert in LINUX or just sharpen your skills.  Sign up for IBM's
> Free Linux Tutorials.  Learn everything from the bash shell to sys admin.
> Click now! http://ads.osdn.com/?ad_id=1278&alloc_id=3371&op=click
> _______________________________________________
> 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: IBM Linux Tutorials.
Become an expert in LINUX or just sharpen your skills.  Sign up for IBM's
Free Linux Tutorials.  Learn everything from the bash shell to sys admin.
Click now! http://ads.osdn.com/?ad_id\x1278&alloc_id371&op=click

Processor C states and USB problem

[mike-Z92qn3yYq0hWk0Htik3J/w]

Recently I had problems with my network card and ACPI (see earlier posts for 
the problem and solution) so I thought I'd take a closer look at ACPI on 
this notebook (Acer Travelmate 800). 

Looking at the proc entries for the processor I saw the following. 

active state:            C3
default state:           C1
bus master activity:     80070020
states:
   C1:                  promotion[C2] demotion[--] latency[000] 
usage[00000010]
   C2:                  promotion[C3] demotion[C1] 
latency[001]usage[01077887]
  *C3:                  promotion[--] demotion[C2] latency[085] 
usage[00083507] 

This looked pretty normal to me, apart from the fact that the processor was 
spending a lot of time in C3 although the notebook was idle. 

Now I plugged in my USB mouse. 

cat /proc/acpi/processor/CPU0/power
active state:            C2
default state:           C1
bus master activity:     ffffffff <---- USB is active
states:
   C1:                  promotion[C2] demotion[--] latency[000] 
usage[00000010]
  *C2:                  promotion[C3] demotion[C1] 
latency[001]usage[00871647]
   C3:                  promotion[--] demotion[C2] latency[085] 
usage[00077386] 

The processor never left the C2 state.
After unplugging the mouse "bus master activity" showed normal values again. 

Now for my two questions: 

1.) Is it normal for the notebook to spend so much time in C3, even if it is 
idle? 

2.) Why is an USB device screwing up the bus master activity, resulting in 
the processor never leaving C2? 


/Michael 




-------------------------------------------------------
This SF.net email is sponsored by: IBM Linux Tutorials.
Become an expert in LINUX or just sharpen your skills.  Sign up for IBM's
Free Linux Tutorials.  Learn everything from the bash shell to sys admin.
Click now! http://ads.osdn.com/?ad_id=1278&alloc_id=3371&op=click