Machine no longer enters C3 on 2.6.11-rc1-bk

Machine no longer enters C3 on 2.6.11-rc1-bk

[Pavel Machek]

Hi!

bus master activity: is changing all the time, mostly 555555555 and
aaaaaaa, and cpu refuses to enter C3 for obvious reasons. I compiled
out USB and sound... Does anybody else see the same problem?

								Pavel
-- 
People were complaining that M$ turns users into beta-testers...
...jr ghea gurz vagb qrirybcref, naq gurl frrz gb yvxr vg gung jnl!

Machine no longer enters C3 on 2.6.11-rc1-bk

[Fred Labrosse]

Pavel Machek writes:
 > Hi!
 > 
 > bus master activity: is changing all the time, mostly 555555555 and
 > aaaaaaa, and cpu refuses to enter C3 for obvious reasons. I compiled
 > out USB and sound... Does anybody else see the same problem?
 > 

I did notice too that it was never in C3 apart from a bit right at the
beginning, since 2.6.10.  It has always been like that up to 2.6.6, became
better from 2.6.7 (but had other problems with that).

Fred

P.S.  I didn't check bus master activity.  How do you do that?





-------------------------------------------------------
This SF.Net email is sponsored by: IntelliVIEW -- Interactive Reporting
Tool for open source databases. Create drag-&-drop reports. Save time
by over 75%! Publish reports on the web. Export to DOC, XLS, RTF, etc.
Download a FREE copy at http://www.intelliview.com/go/osdn_nl

Re: Machine no longer enters C3 on 2.6.11-rc1-bk

[Pavel Machek]

Hi!

>  > bus master activity: is changing all the time, mostly 555555555 and
>  > aaaaaaa, and cpu refuses to enter C3 for obvious reasons. I compiled
>  > out USB and sound... Does anybody else see the same problem?
>  > 
> 
> I did notice too that it was never in C3 apart from a bit right at the
> beginning, since 2.6.10.  It has always been like that up to 2.6.6, became
> better from 2.6.7 (but had other problems with that).

Good, so I'm not alone.

> P.S.  I didn't check bus master activity.  How do you do that?

cat /proc/acpi/processor/*/power

								Pavel

-- 
People were complaining that M$ turns users into beta-testers...
...jr ghea gurz vagb qrirybcref, naq gurl frrz gb yvxr vg gung jnl!


-------------------------------------------------------
This SF.Net email is sponsored by: IntelliVIEW -- Interactive Reporting
Tool for open source databases. Create drag-&-drop reports. Save time
by over 75%! Publish reports on the web. Export to DOC, XLS, RTF, etc.
Download a FREE copy at http://www.intelliview.com/go/osdn_nl

Re: Machine no longer enters C3 on 2.6.11-rc1-bk

[Rich Townsend]

Pavel Machek wrote:
> Hi!
> 
> 
>> > bus master activity: is changing all the time, mostly 555555555 and
>> > aaaaaaa, and cpu refuses to enter C3 for obvious reasons. I compiled
>> > out USB and sound... Does anybody else see the same problem?
>> > 
>>
>>I did notice too that it was never in C3 apart from a bit right at the
>>beginning, since 2.6.10.  It has always been like that up to 2.6.6, became
>>better from 2.6.7 (but had other problems with that).
> 
> 
> Good, so I'm not alone.
> 

I see the same thing. Furthermore, my CPU never appears to enter C1 
either (except at the beginning). Is this aspect of the behaviour normal?

By the way, the bus master activity always appears as ffffffff.

cheers,

Rich


-------------------------------------------------------
This SF.Net email is sponsored by: IntelliVIEW -- Interactive Reporting
Tool for open source databases. Create drag-&-drop reports. Save time
by over 75%! Publish reports on the web. Export to DOC, XLS, RTF, etc.
Download a FREE copy at http://www.intelliview.com/go/osdn_nl

Re: Machine no longer enters C3 on 2.6.11-rc1-bk

[Rich Townsend]

Rich Townsend wrote:
> Pavel Machek wrote:
> 
>> Hi!
>>
>>
>>> > bus master activity: is changing all the time, mostly 555555555 and
>>> > aaaaaaa, and cpu refuses to enter C3 for obvious reasons. I compiled
>>> > out USB and sound... Does anybody else see the same problem?
>>> >
>>> I did notice too that it was never in C3 apart from a bit right at the
>>> beginning, since 2.6.10.  It has always been like that up to 2.6.6, 
>>> became
>>> better from 2.6.7 (but had other problems with that).
>>
>>
>>
>> Good, so I'm not alone.
>>
> 
> I see the same thing. Furthermore, my CPU never appears to enter C1 
> either (except at the beginning). Is this aspect of the behaviour normal?
> 
> By the way, the bus master activity always appears as ffffffff.

Gak, I just read the wiki, and USB was the culprit. But that doesn't 
explain why I'm not seeing any transitions to C1, even under full load.

cheers,

Rich


-------------------------------------------------------
This SF.Net email is sponsored by: IntelliVIEW -- Interactive Reporting
Tool for open source databases. Create drag-&-drop reports. Save time
by over 75%! Publish reports on the web. Export to DOC, XLS, RTF, etc.
Download a FREE copy at http://www.intelliview.com/go/osdn_nl

Re: Machine no longer enters C3 on 2.6.11-rc1-bk

[Len Brown]

On Thu, 2005-01-20 at 09:32, Rich Townsend wrote:

> Gak, I just read the wiki, and USB was the culprit. But that doesn't 
> explain why I'm not seeing any transitions to C1, even under full load.

This is normal for the current code -- we run C1 16 times and promote,
but if C2 latency is very low, we never demote to C1:

[root@d600 root]#  cat  /proc/acpi/processor/CPU0/power
active state:            C4
max_cstate:              C8
bus master activity:     00000000
states:
 C1: type[C1] promotion[C2] demotion[--] latency[001] usage[00000010] 
 C2: type[C2] promotion[C3] demotion[C1] latency[001] usage[00266192]
 C3: type[C3] promotion[C4] demotion[C2] latency[085] usage[00028664]
*C4: type[C3] promotion[--] demotion[C3] latency[185] usage[00359811]




-------------------------------------------------------
This SF.Net email is sponsored by: IntelliVIEW -- Interactive Reporting
Tool for open source databases. Create drag-&-drop reports. Save time
by over 75%! Publish reports on the web. Export to DOC, XLS, RTF, etc.
Download a FREE copy at http://www.intelliview.com/go/osdn_nl

Re: Machine no longer enters C3 on 2.6.11-rc1-bk

[Pavel Machek]

Hi!

>  > bus master activity: is changing all the time, mostly 555555555 and
>  > aaaaaaa, and cpu refuses to enter C3 for obvious reasons. I compiled
>  > out USB and sound... Does anybody else see the same problem?
>  > 
> 
> I did notice too that it was never in C3 apart from a bit right at the
> beginning, since 2.6.10.  It has always been like that up to 2.6.6, became
> better from 2.6.7 (but had other problems with that).
> 
> P.S.  I didn't check bus master activity.  How do you do that?

There's something wrong with bus master activity indicator: it seems
to be updated every HZ or something like that... that results in a bad
stuff if HZ=100... activiy is now non-zero and machine refuses to go
C3.
								Pavel
-- 
People were complaining that M$ turns users into beta-testers...
...jr ghea gurz vagb qrirybcref, naq gurl frrz gb yvxr vg gung jnl!


-------------------------------------------------------
This SF.Net email is sponsored by: IntelliVIEW -- Interactive Reporting
Tool for open source databases. Create drag-&-drop reports. Save time
by over 75%! Publish reports on the web. Export to DOC, XLS, RTF, etc.
Download a FREE copy at http://www.intelliview.com/go/osdn_nl

Re: Machine no longer enters C3 on 2.6.11-rc1-bk

[Pavel Machek]

Hi!

>  > bus master activity: is changing all the time, mostly 555555555 and
>  > aaaaaaa, and cpu refuses to enter C3 for obvious reasons. I compiled
>  > out USB and sound... Does anybody else see the same problem?
>  > 
> 
> I did notice too that it was never in C3 apart from a bit right at the
> beginning, since 2.6.10.  It has always been like that up to 2.6.6, became
> better from 2.6.7 (but had other problems with that).
> 
> Fred
> 
> P.S.  I didn't check bus master activity.  How do you do that?

can you try to kill all bm_activity++ things in
drivers/acpi/processor.c and see what it does to your power
consumption. (Warning: it may do bad stuff to your performance).

								Pavel
-- 
People were complaining that M$ turns users into beta-testers...
...jr ghea gurz vagb qrirybcref, naq gurl frrz gb yvxr vg gung jnl!


-------------------------------------------------------
This SF.Net email is sponsored by: IntelliVIEW -- Interactive Reporting
Tool for open source databases. Create drag-&-drop reports. Save time
by over 75%! Publish reports on the web. Export to DOC, XLS, RTF, etc.
Download a FREE copy at http://www.intelliview.com/go/osdn_nl