Re: BIOS and CPU C_states are strange

[yakui_zhao <yakui.zhao@intel.com>]

On Sun, 2009-06-21 at 17:20 +0800, Mahmood Naderan wrote:
> I got more confused... In my first post if you see, BIOS reports C1,
> C2 and C6. Even it does not support C3 (!),  but here is the output of
> "cat /proc/acpi/processor/CPU0/power":
> 
> active state:            C0
> max_cstate:              C8
> bus master activity:     00000000
> maximum allowed latency: 2000000000 usec
> states:
>    C1:                  type[C1] promotion[--] demotion[--]
> latency[001] usage[00000063] duration[00000000000000000000]
>    C2:                  type[C2] promotion[--] demotion[--]
> latency[001] usage[00017466] duration[00000000000014422985]
>    C3:                  type[C3] promotion[--] demotion[--]
> latency[162] usage[00096196] duration[00000000000650582389]
> 
> So I think something (BIOS or CPU or OS) is malfunctioning and does
> not report correctly.
> Now the big question is, will my system (a combination of CPU, BIOS
> and OS) enter C3 (or even deeper states)?
What is your problem?

The following info only tells us that the following CPU C-state is
supported on the CPU.
>Your CPU supports the following C-states : C1 C2 C3 C4 C5 C6

But not all the CPU C-state is not used by the OS.
It seems that the C1/C2/C6 is used on this box.

At the same time it is noted that the cpu C-state is different with the
ACPI C-state. In fact BIOS will do the mapping between the CPU C-state
and ACPI C-state. And ACPI C-state is used by the OS.
For example: on your box: The CPU C6 is mapped to ACPI C3-state.

>From the info of "cat /proc/acpi/processor/CPU*/power" we know that your
box can enter the ACPI C3, which is mapped to the CPU C6. Of course it
is the deep C-state.

Thanks.

> 
> --------------------------
> *Mahmood Naderan*
> 
> 
> 
> 
> On Sun, Jun 21, 2009 at 2:34 AM, Ananth Narayan
> S<ananth.narayan@gmail.com> wrote:
> > You could see it that way. The OS will use the C states exported by
> > the BIOS. If C4/C5 are not exported, the OS will not even know that
> > such states exist. If you query /proc/acpi/CPU*/power, you'll probably
> > see just three C states listed (C1, C2, C3). Typically the last
> > C-state (C3) in that will map to the lowest C-state supported by the
> > processor.
> >
> > -- Ananth
> >
> >
> >
> >
> > On Sat, Jun 20, 2009 at 8:46 AM, Mahmood Naderan<mahmood.nt@gmail.com> wrote:
> >> Hi,
> >> So if my understanding is correct, the BIOS does not allow the CPU to
> >> enter C4 and C5 states. Right?
> >>
> >> --------------------------
> >> *Mahmood Naderan*
> >>
> >>
> >>
> >>
> >> On Sat, Jun 20, 2009 at 7:56 PM, Ananth Narayan
> >> S<ananth.narayan@gmail.com> wrote:
> >>> Intel processors export a max supported c state value. But when it
> >>> comes to c states, typically the one that provides max power savings
> >>> is exported by the BIOS. The intermediate ones aren't.
> >>>
> >>> -- Ananth Narayan S.
> >>>
> >>>
> >>>
> >>>
> >>> On Fri, Jun 19, 2009 at 10:07 AM, Mahmood Naderan<mahmood.nt@gmail.com> wrote:
> >>>> Hi,
> >>>> I have run powertop with root permission with both AC and batter powers. The
> >>>> reported C_States are the same and strange:
> >>>>
> >>>> mahmood@magma:~$ sudo powertop
> >>>> PowerTOP 1.11   (C) 2007, 2008 Intel Corporation
> >>>>
> >>>> Collecting data for 5 seconds
> >>>>
> >>>> Your CPU supports the following C-states : C1 C2 C3 C4 C5 C6
> >>>> Your BIOS reports the following C-states : C1 C2 C6
> >>>>
> >>>> Where are C4 and C5?
> >>>>
> >>>> --------------------------
> >>>> *Mahmood Naderan*
> >>>> --
> >>>> To unsubscribe from this list: send the line "unsubscribe linux-acpi" in
> >>>> the body of a message to majordomo@vger.kernel.org
> >>>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> >>>>
> >>>
> >>
> >
> --
> To unsubscribe from this list: send the line "unsubscribe linux-acpi" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

--
To unsubscribe from this list: send the line "unsubscribe linux-acpi" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html