poll mode is more lazy to speed down, is this an idle system?(what's platform?)
if so, you should wait until the frequency speed down to the lowest one and
read the power usage again.

It looks like xcall/pstate transition is gone. I suspect latest scheduler 
change makes
the system more aggressive to race to idle.

I attached a dtrace script to capture the idle/non-idle transition number. This 
is not
p/c-state related, and not powertop related. I saw a big regression against 
previous
build. Please have a try on your system and let me know the result.

Thanks,
-Aubrey

John Martin wrote:
> 
> On 04/ 5/11 10:37 AM, Aubrey Li wrote:
> 
> >> If I change the last line to:
> >>
> >>   cpupm  enable event-mode
> >
> >
> > typo??
> >
> > it should be
> > cpupm enable poll-mode
> 
> Sorry, I thought the default was poll-mode.  If I change it to:
> 
>    cpupm  enable poll-mode
> 
> after running pmconfig, the average wall outlet power use goes
> up 10-20W with peaks 40W higher.  Powertop output:
> 
> 
> C-states (idle power) Avg     Residency                       P-states 
> (frequencies)
> C0 (cpu       running)                (14.0%)                         1596 
> Mhz        0.0%
> C1                    2.0ms   (39.0%)                         1729 Mhz        
> 0.0%
> C2                    0.0ms   (0.0%)                          1862 Mhz        
> 0.0%
> C3                    0.0ms   (47.0%)                         1995 Mhz        
> 0.0%
> 
> 2128
> Mhz   0.0%
> 
> 2261
> Mhz   0.0%
> 
> 2394
> Mhz   0.0%
> 
> 2527
> Mhz   0.0%
> 
> 2660
> Mhz   0.0%
> 
> 2793
> Mhz   0.0%
> 
> 2926
> Mhz   0.0%
> 
> 3059
> Mhz   0.0%
> 
> 3192
> Mhz   0.0%
> 
> 3325
> Mhz   12.4%
> 
> 3457
> Mhz(turbo)    87.6%
> 
> Wakeups-from-idle per second: 270972.6  interval: 5.0s
> no ACPI       power usage estimate available
> 
> Top causes for wakeups:
> 10.6% (28636.6)                sched :        <xcalls> unix`dtrace_xcall_func
>   0.3% (702.4)                  sched :       <xcalls>
> unix`hati_demap_func
>   0.2% (609.0)               <kernel> :       genunix`clock
>   0.2% (590.6)               <kernel> :       genunix`cv_wakeup
>   0.1% (304.6)               <kernel> :       SDC`sysdc_update
>   0.1% (229.1)               <kernel> :       genunix`lwp_timer_timeout
>   0.1% (140.5)               <kernel> :
> uhci`uhci_handle_root_hub_status_change
>   0.0% (      73.9)               <kernel> :  cpudrv`cpudrv_monitor_disp
>   0.0% (      71.9)               <kernel> :  genunix`pm_rescan
>   0.0% (      68.9)            <interrupt> :  ehci#0
>   0.0% (      47.1)               <kernel> :
> ehci`ehci_handle_root_hub_status_change
>   0.0% (      36.3)               <kernel> :  uhci`uhci_cmd_timeout_hdlr
>   0.0% (      25.1)            <interrupt> :  ahci#0
>   0.0% (      24.4)               <kernel> :  genunix`schedpaging
>   0.0% (      15.8)                  sched :  <xcalls>
> unix`speedstep_pstate_transition
>   0.0% (       6.8)            <interrupt> :  nvidia#0
>   0.0% (       6.2)               <kernel> :  nvidia`nvidia_rc_timer
>   0.0% (       6.0)            <interrupt> :  uhci#2
>   0.0% (       6.0)            <interrupt> :  uhci#0
>   0.0% (       6.0)               <kernel> :  TS`ts_update
>   0.0% (       6.0)            <interrupt> :  uhci#4
>   0.0% (       6.0)            <interrupt> :  uhci#3

Attachment: idle-state.d
Description: idle-state.d

_______________________________________________
pm-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/pm-discuss

Reply via email to